Tool help documentation

Each of the following functions are methods of WbEnvironment class. Tools may be called using the convention in the following example:

import whitebox_workflows as wbw

wbe = wbw.WbEnvironment()
# Set up the environment, e.g. working directory, verbose mode, max_procs
raster = wbe.read_raster('my_raster.tif') # Read some kind of data
result = wbe.mean_filter(raster) # Call some kind of function
...
  1. adaptive_filter
  2. add_point_coordinates_to_table
  3. aggregate_raster
  4. anova
  5. ascii_to_las
  6. aspect
  7. attribute_correlation
  8. attribute_histogram
  9. attribute_scattergram
  10. average_flowpath_slope
  11. average_normal_vector_angular_deviation
  12. average_overlay
  13. average_upslope_flowpath_length
  14. balance_contrast_enhancement
  15. basins
  16. bilateral_filter
  17. block_maximum
  18. block_minimum
  19. bool_and
  20. bool_not
  21. bool_or
  22. bool_xor
  23. boundary_shape_complexity
  24. breach_depressions_least_cost
  25. breach_single_cell_pits
  26. buffer_raster
  27. burn_streams_at_roads
  28. centroid_raster
  29. centroid_vector
  30. change_vector_analysis
  31. circular_variance_of_aspect
  32. classify_buildings_in_lidar
  33. classify_overlap_points
  34. clean_vector
  35. clip
  36. clip_lidar_to_polygon
  37. clip_raster_to_polygon
  38. closing
  39. clump
  40. compactness_ratio
  41. conservative_smoothing_filter
  42. construct_vector_tin
  43. contours_from_points
  44. contours_from_raster
  45. convert_nodata_to_zero
  46. corner_detection
  47. correct_vignetting
  48. cost_allocation
  49. cost_distance
  50. cost_pathway
  51. count_if
  52. create_colour_composite
  53. create_plane
  54. crispness_index
  55. cross_tabulation
  56. csv_points_to_vector
  57. cumulative_distribution
  58. d8_flow_accum
  59. d8_mass_flux
  60. d8_pointer
  61. depth_in_sink
  62. deviation_from_mean_elevation
  63. diff_of_gaussians_filter
  64. difference
  65. difference_from_mean_elevation
  66. dinf_flow_accum
  67. dinf_mass_flux
  68. dinf_pointer
  69. direct_decorrelation_stretch
  70. directional_relief
  71. dissolve
  72. distance_to_outlet
  73. diversity_filter
  74. downslope_distance_to_stream
  75. downslope_flowpath_length
  76. downslope_index
  77. edge_contamination
  78. edge_density
  79. edge_preserving_mean_filter
  80. edge_proportion
  81. elev_relative_to_min_max
  82. elev_relative_to_watershed_min_max
  83. elevation_above_pit
  84. elevation_above_stream
  85. elevation_above_stream_euclidean
  86. elevation_percentile
  87. eliminate_coincident_points
  88. elongation_ratio
  89. embankment_mapping
  90. emboss_filter
  91. erase
  92. erase_polygon_from_lidar
  93. erase_polygon_from_raster
  94. euclidean_allocation
  95. euclidean_distance
  96. export_table_to_csv
  97. exposure_towards_wind_flux
  98. extend_vector_lines
  99. extract_nodes
  100. extract_raster_values_at_points
  101. extract_streams
  102. extract_valleys
  103. farthest_channel_head
  104. fast_almost_gaussian_filter
  105. fd8_flow_accum
  106. fd8_pointer
  107. feature_preserving_smoothing
  108. fetch_analysis
  109. fill_burn
  110. fill_depressions
  111. fill_depressions_planchon_and_darboux
  112. fill_depressions_wang_and_liu
  113. fill_missing_data
  114. fill_pits
  115. filter_lidar_classes
  116. filter_lidar_scan_angles
  117. filter_raster_features_by_area
  118. find_flightline_edge_points
  119. find_lowest_or_highest_points
  120. find_main_stem
  121. find_noflow_cells
  122. find_parallel_flow
  123. find_patch_edge_cells
  124. find_ridges
  125. flatten_lakes
  126. flightline_overlap
  127. flip_image
  128. flood_order
  129. flow_accum_full_workflow
  130. flow_length_diff
  131. gamma_correction
  132. gaussian_contrast_stretch
  133. gaussian_curvature
  134. gaussian_filter
  135. geomorphons
  136. hack_stream_order
  137. heat_map
  138. height_above_ground
  139. hexagonal_grid_from_raster_base
  140. hexagonal_grid_from_vector_base
  141. high_pass_filter
  142. high_pass_median_filter
  143. highest_position
  144. hillshade
  145. hillslopes
  146. histogram_equalization
  147. histogram_matching
  148. histogram_matching_two_images
  149. hole_proportion
  150. horizon_angle
  151. horton_stream_order
  152. hypsometric_analysis
  153. hypsometrically_tinted_hillshade
  154. idw_interpolation
  155. ihs_to_rgb
  156. image_autocorrelation
  157. image_correlation
  158. image_correlation_neighbourhood_analysis
  159. image_regression
  160. image_stack_profile
  161. impoundment_size_index
  162. insert_dams
  163. integral_image_transform
  164. intersect
  165. isobasins
  166. jenson_snap_pour_points
  167. join_tables
  168. k_means_clustering
  169. k_nearest_mean_filter
  170. kappa_index
  171. ks_normality_test
  172. laplacian_filter
  173. laplacian_of_gaussians_filter
  174. las_to_ascii
  175. las_to_shapefile
  176. layer_footprint_raster
  177. layer_footprint_vector
  178. lee_filter
  179. length_of_upstream_channels
  180. lidar_block_maximum
  181. lidar_block_minimum
  182. lidar_classify_subset
  183. lidar_colourize
  184. lidar_construct_vector_tin
  185. lidar_digital_surface_model
  186. lidar_elevation_slice
  187. lidar_ground_point_filter
  188. lidar_hex_bin
  189. lidar_hillshade
  190. lidar_histogram
  191. lidar_idw_interpolation
  192. lidar_info
  193. lidar_join
  194. lidar_kappa
  195. lidar_nearest_neighbour_gridding
  196. lidar_point_density
  197. lidar_point_stats
  198. lidar_radial_basis_function_interpolation
  199. lidar_ransac_planes
  200. lidar_remove_outliers
  201. lidar_rooftop_analysis
  202. lidar_segmentation
  203. lidar_segmentation_based_filter
  204. lidar_shift
  205. lidar_thin
  206. lidar_thin_high_density
  207. lidar_tile
  208. lidar_tile_footprint
  209. lidar_tin_gridding
  210. lidar_tophat_transform
  211. line_detection_filter
  212. line_intersections
  213. line_thinning
  214. linearity_index
  215. lines_to_polygons
  216. list_unique_values
  217. long_profile
  218. long_profile_from_points
  219. longest_flowpath
  220. lowest_position
  221. majority_filter
  222. map_off_terrain_objects
  223. max_absolute_overlay
  224. max_anisotropy_dev
  225. max_anisotropy_dev_signature
  226. max_branch_length
  227. max_difference_from_mean
  228. max_downslope_elev_change
  229. max_elevation_dev_signature
  230. max_elevation_deviation
  231. max_overlay
  232. max_procs
  233. max_upslope_elev_change
  234. max_upslope_flowpath_length
  235. max_upslope_value
  236. maximal_curvature
  237. maximum_filter
  238. mdinf_flow_accum
  239. mean_curvature
  240. mean_filter
  241. median_filter
  242. medoid
  243. merge_line_segments
  244. merge_table_with_csv
  245. merge_vectors
  246. min_absolute_overlay
  247. min_downslope_elev_change
  248. min_max_contrast_stretch
  249. min_overlay
  250. minimal_curvature
  251. minimum_bounding_box
  252. minimum_bounding_circle
  253. minimum_bounding_envelope
  254. minimum_convex_hull
  255. minimum_filter
  256. modified_k_means_clustering
  257. modify_nodata_value
  258. mosaic
  259. mosaic_with_feathering
  260. multidirectional_hillshade
  261. multipart_to_singlepart
  262. multiply_overlay
  263. multiscale_elevation_percentile
  264. multiscale_roughness
  265. multiscale_roughness_signature
  266. multiscale_std_dev_normals
  267. multiscale_std_dev_normals_signature
  268. multiscale_topographic_position_image
  269. narrowness_index
  270. natural_neighbour_interpolation
  271. nearest_neighbour_interpolation
  272. new_lidar
  273. new_raster
  274. new_raster_from_base_raster
  275. new_raster_from_base_vector
  276. new_vector
  277. normal_vectors
  278. normalized_difference_index
  279. num_downslope_neighbours
  280. num_inflowing_neighbours
  281. olympic_filter
  282. opening
  283. paired_sample_t_test
  284. panchromatic_sharpening
  285. patch_orientation
  286. pennock_landform_classification
  287. percent_elev_range
  288. percent_equal_to
  289. percent_greater_than
  290. percent_less_than
  291. percentage_contrast_stretch
  292. percentile_filter
  293. perimeter_area_ratio
  294. pick_from_list
  295. plan_curvature
  296. polygon_area
  297. polygon_long_axis
  298. polygon_perimeter
  299. polygon_short_axis
  300. polygonize
  301. polygons_to_lines
  302. prewitt_filter
  303. principal_component_analysis
  304. print_geotiff_tags
  305. profile
  306. profile_curvature
  307. qin_flow_accumulation
  308. quantiles
  309. quinn_flow_accumulation
  310. radial_basis_function_interpolation
  311. radius_of_gyration
  312. raise_walls
  313. random_field
  314. random_sample
  315. range_filter
  316. raster_area
  317. raster_calculator
  318. raster_cell_assignment
  319. raster_histogram
  320. raster_perimeter
  321. raster_streams_to_vector
  322. raster_summary_stats
  323. raster_to_vector_lines
  324. raster_to_vector_points
  325. raster_to_vector_polygons
  326. rasterize_streams
  327. read_lidar
  328. read_lidars
  329. read_raster
  330. read_rasters
  331. read_vector
  332. read_vectors
  333. reciprocal
  334. reclass
  335. reclass_equal_interval
  336. rectangular_grid_from_raster_base
  337. rectangular_grid_from_vector_base
  338. reinitialize_attribute_table
  339. related_circumscribing_circle
  340. relative_aspect
  341. relative_stream_power_index
  342. relative_topographic_position
  343. remove_duplicates
  344. remove_off_terrain_objects
  345. remove_polygon_holes
  346. remove_short_streams
  347. remove_spurs
  348. resample
  349. rescale_value_range
  350. rgb_to_ihs
  351. rho8_flow_accum
  352. rho8_pointer
  353. roberts_cross_filter
  354. root_mean_square_error
  355. ruggedness_index
  356. scharr_filter
  357. sediment_transport_index
  358. select_tiles_by_polygon
  359. set_nodata_value
  360. shape_complexity_index_raster
  361. shape_complexity_index_vector
  362. shreve_stream_magnitude
  363. sigmoidal_contrast_stretch
  364. singlepart_to_multipart
  365. sink
  366. slope
  367. slope_vs_elev_plot
  368. smooth_vectors
  369. snap_pour_points
  370. sobel_filter
  371. spherical_std_dev_of_normals
  372. split_colour_composite
  373. split_vector_lines
  374. split_with_lines
  375. standard_deviation_contrast_stretch
  376. standard_deviation_filter
  377. standard_deviation_of_slope
  378. stochastic_depression_analysis
  379. strahler_order_basins
  380. strahler_stream_order
  381. stream_link_class
  382. stream_link_identifier
  383. stream_link_length
  384. stream_link_slope
  385. stream_slope_continuous
  386. subbasins
  387. sum_overlay
  388. surface_area_ratio
  389. symmetrical_difference
  390. tangential_curvature
  391. thicken_raster_line
  392. time_in_daylight
  393. tin_interpolation
  394. tophat_transform
  395. topological_stream_order
  396. total_curvature
  397. total_filter
  398. trace_downslope_flowpaths
  399. travelling_salesman_problem
  400. trend_surface
  401. trend_surface_vector_points
  402. tributary_identifier
  403. turning_bands_simulation
  404. two_sample_ks_test
  405. union
  406. unnest_basins
  407. unsharp_masking
  408. update_nodata_cells
  409. upslope_depression_storage
  410. user_defined_weights_filter
  411. vector_hex_binning
  412. vector_lines_to_raster
  413. vector_points_to_raster
  414. vector_polygons_to_raster
  415. verbose
  416. version
  417. viewshed
  418. visibility_index
  419. voronoi_diagram
  420. watershed
  421. watershed_from_raster_pour_points
  422. weighted_overlay
  423. weighted_sum
  424. wetness_index
  425. wilcoxon_signed_rank_test
  426. working_directory
  427. write_function_memory_insertion
  428. write_lidar
  429. write_raster
  430. write_vector
  431. z_scores
  432. zonal_statistics

adaptive_filter

This tool performs a type of adaptive filter on a raster image. An adaptive filter can be used to reduce the level of random noise (shot noise) in an image. The algorithm operates by calculating the average value in a moving window centred on each grid cell. If the absolute difference between the window mean value and the centre grid cell value is beyond a user-defined threshold (threshold), the grid cell in the output image is assigned the mean value, otherwise it is equivalent to the original value. Therefore, the algorithm only modifies the image where grid cell values are substantially different than their neighbouring values.

Neighbourhood size, or filter size, is specified in the x and y dimensions using filterx and filtery. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

See Also

mean_filter

Function Signature

def adaptive_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11, threshold: float = 2.0) -> Raster: ...

add_point_coordinates_to_table

Description

This tool modifies the attribute table of a vector of POINT VectorGeometryType by adding two fields, XCOORD and YCOORD, containing each point's X and Y coordinates respectively.

Parameters

input (Vector): The input Vector object

Returns

Vector: the returning value

Function Signature

def add_point_coordinates_to_table(self, input: Vector) -> Vector: ...

aggregate_raster

This tool can be used to reduce the grid resolution of a raster by a user specified amount. For example, using an aggregation factor (agg_factor) of 2 would result in a raster with half the number of rows and columns. The grid cell values (type) in the output image will consist of the mean, sum, maximum, minimum, or range of the overlapping grid cells in the input raster (four cells in the case of an aggregation factor of 2).

See Also

resample

Function Signature

def aggregate_raster(self, raster: Raster, aggregation_factor: int = 2, aggregation_type: str = "mean") -> Raster: ...

anova

This tool performs an Analysis of variance (ANOVA) test on the distribution of values in a raster (input) among a group of features (features). The ANOVA report is written to an output HTML report (output).

Function Signature

def anova(self, input_raster: Raster, features_raster: Raster, output_html_file: str) -> None: ...

ascii_to_las

This tool can be used to convert one or more ASCII files, containing LiDAR point data, into LAS files. The user must specify the name(s) of the input ASCII file(s) (inputs). Each input file will have a correspondingly named output file with a .las file extension. The output point data, each on a separate line, will take the format:

x,y,z,intensity,class,return,num_returns"
ValueInterpretation
xx-coordinate
yy-coordinate
zelevation
iintensity value
cclassification
rnreturn number
nrnumber of returns
timeGPS time
sascan angle
rred
bblue
ggreen

The x, y, and z patterns must always be specified. If the rn pattern is used, the nr pattern must also be specified. Examples of valid pattern string include:

'x,y,z,i'
'x,y,z,i,rn,nr'
'x,y,z,i,c,rn,nr,sa'
'z,x,y,rn,nr'
'x,y,z,i,rn,nr,r,g,b'

Use the las_to_ascii tool to convert a LAS file into a text file containing LiDAR point data.

See Also

las_to_ascii

Function Signature

def ascii_to_las(self, input_ascii_files: List[str], pattern: str, epsg_code: int) -> None: ...

aspect

This tool calculates slope aspect (i.e. slope orientation in degrees clockwise from north) for each grid cell in an input digital elevation model (DEM). The user must specify an input DEM (dem). The Z conversion factor is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z conversion factor. If the DEM is in the geographic coordinate system (latitude and longitude), the following equation is used:

zfactor = 1.0 / (111320.0 x cos(mid_lat))

where mid_lat is the latitude of the centre of each raster row, in radians.

The tool uses Horn's (1981) 3rd-order finite difference method to estimate slope. Given the following clock-type grid cell numbering scheme (Gallant and Wilson, 2000),

| 7 | 8 | 1 |
| 6 | 9 | 2 |
| 5 | 4 | 3 |

aspect = 180 - arctan(fy / fx) + 90(fx / |fx|)

where,

fx = (z3 - z5 + 2(z2 - z6) + z1 - z7) / 8 * Δx

and,

fy = (z7 - z5 + 2(z8 - z4) + z1 - z3) / 8 * Δy

Δx and Δy are the grid resolutions in the x and y direction respectively

Reference

Gallant, J. C., and J. P. Wilson, 2000, Primary topographic attributes, in Terrain Analysis: Principles and Applications, edited by J. P. Wilson and J. C. Gallant pp. 51-86, John Wiley, Hoboken, N.J.

See Also

slope, plan_curvature, profile_curvature

Function Signature

def aspect(self, dem: Raster, z_factor: float = 1.0) -> Raster: ...

attribute_correlation

This tool can be used to estimate the Pearson product-moment correlation coefficient (r) for each pair among a group of attributes associated with the database file of a shapefile. The r-value is a measure of the linear association in the variation of the attributes. The coefficient ranges from -1, indicated a perfect negative linear association, to 1, indicated a perfect positive linear association. An r-value of 0 indicates no correlation between the test variables.

Notice that this index is a measure of the linear association; two variables may be strongly related by a non-linear association (e.g. a power function curve) which will lead to an apparent weak association based on the Pearson coefficient. In fact, non-linear associations are very common among spatial variables, e.g. terrain indices such as slope and contributing area. In such cases, it is advisable that the input images are transformed prior to the estimation of the Pearson coefficient, or that an alternative, non-parametric statistic be used, e.g. the Spearman rank correlation coefficient.

The user must specify the name of the input vector Shapefile (input). Correlations will be calculated for each pair of numerical attributes contained within the input file's attribute table and presented in a correlation matrix HMTL output (output).

See Also

image_correlation, attribute_scattergram, attribute_histogram

Function Signature

def attribute_correlation(self, input: Vector, output_html_file: str) -> None: ...

attribute_histogram

This tool can be used to create a histogram, which is a graph displaying the frequency distribution of data, for the values contained in a field of an input vector's attribute table. The user must specify the name of an input vector (input) and the name of one of the fields (field) contained in the associated attribute table. The tool output (output) is an HTML formatted histogram analysis report. If the specified field is non-numerical, the tool will produce a bar-chart of class frequency, similar to the tabular output of the list_unique_values tool.

See Also

list_unique_values, raster_histogram

Function Signature

def attribute_histogram(self, input: Vector, field_name: str, output_html_file: str) -> None: ...

attribute_scattergram

This tool can be used to create a scattergram for two numerical fields (fieldx and fieldy) contained within an input vector's attribute table (input). The user must specify the name of an input shapefile and the name of two of the fields contained it the associated attribute table. The tool output (output) is an HTML formatted report containing a graphical scattergram plot.

See Also

attribute_histogram, attribute_correlation

Function Signature

def attribute_scattergram(self, input: Vector, field_name_x: str, field_name_y: str, output_html_file: str, add_trendline: bool = False) -> None: ...

average_flowpath_slope

This tool calculates the average slope gradient (i.e. slope steepness in degrees) of the flowpaths that pass through each grid cell in an input digital elevation model (DEM). The user must specify the name of a DEM raster (dem). It is important that this DEM is pre-processed to remove all topographic depressions and flat areas using a tool such as breach_depressions_least_cost. Several intermediate rasters are created and stored in memory during the operation of this tool, which may limit the size of DEM that can be processed, depending on available system resources.

See Also

average_upslope_flowpath_length, breach_depressions_least_cost

Function Signature

def average_flowpath_slope(self, dem: Raster) -> Raster: ...

average_normal_vector_angular_deviation

This tool characterizes the spatial distribution of the average normal vector angular deviation, a measure of surface roughness. Working in the field of 3D printing, Ko et al. (2016) defined a measure of surface roughness based on quantifying the angular deviations in the direction of the normal vector of a real surface from its ideal (i.e. smoothed) form. This measure of surface complexity is therefore in units of degrees. Specifically, roughness is defined in this study as the neighborhood-averaged difference in the normal vectors of the original DEM and a smoothed DEM surface. Smoothed surfaces are derived by applying a Gaussian blur of the same size as the neighborhood (filter).

The multiscale_roughness tool calculates the same measure of surface roughness, except that it is designed to work with multiple spatial scales.

Reference

Ko, M., Kang, H., ulrim Kim, J., Lee, Y., & Hwang, J. E. (2016, July). How to measure quality of affordable 3D printing: Cultivating quantitative index in the user community. In International Conference on Human-Computer Interaction (pp. 116-121). Springer, Cham.

Lindsay, J. B., & Newman, D. R. (2018). Hyper-scale analysis of surface roughness. PeerJ Preprints, 6, e27110v1.

See Also

multiscale_roughness, spherical_std_dev_of_normals, circular_variance_of_aspect

Function Signature

def average_normal_vector_angular_deviation(self, dem: Raster, filter_size: int = 11) -> Raster: ...

average_overlay

This tool can be used to find the average value in each cell of a grid from a set of input images (inputs). It is therefore similar to the weighted_sum tool except that each input image is given equal weighting. This tool operates on a cell-by-cell basis. Therefore, each of the input rasters must share the same number of rows and columns and spatial extent. An error will be issued if this is not the case. At least two input rasters are required to run this tool. Like each of the WhiteboxTools overlay tools, this tool has been optimized for parallel processing.

See Also

weighted_sum

Function Signature

def average_overlay(self, input_rasters: List[Raster]) -> Raster: ...

average_upslope_flowpath_length

This tool calculates the average slope gradient (i.e. slope steepness in degrees) of the flowpaths that pass through each grid cell in an input digital elevation model (DEM). The user must specify the name of a DEM raster (dem). It is important that this DEM is pre-processed to remove all topographic depressions and flat areas using a tool such as breach_depressions_least_cost. Several intermediate rasters are created and stored in memory during the operation of this tool, which may limit the size of DEM that can be processed, depending on available system resources.

See Also

average_upslope_flowpath_length, breach_depressions_least_cost

Function Signature

def average_upslope_flowpath_length(self, dem: Raster) -> Raster: ...

balance_contrast_enhancement

This tool can be used to reduce colour bias in a colour composite image based on the technique described by Liu (1991). Colour bias is a common phenomena with colour images derived from multispectral imagery, whereby a higher average brightness value in one band results in over-representation of that band in the colour composite. The tool essentially applies a parabolic stretch to each of the three bands in a user specified RGB colour composite, forcing the histograms of each band to have the same minimum, maximum, and average values while maintaining their overall histogram shape. For greater detail on the operation of the tool, please see Liu (1991). Aside from the names of the input and output colour composite images, the user must also set the value of E, the desired output band mean, where 20 < E < 235.

Reference

Liu, J.G. (1991) Balance contrast enhancement technique and its application in image colour composition. International Journal of Remote Sensing, 12:10.

See Also

direct_decorrelation_stretch, histogram_matching, histogram_matching_two_images, histogram_equalization, gaussian_contrast_stretch

Function Signature

def balance_contrast_enhancement(self, image: Raster, band_mean: float = 100.0) -> Raster: ...

basins

This tool can be used to delineate all of the drainage basins contained within a local drainage direction, or flow pointer raster (d8_pntr), and draining to the edge of the data. The flow pointer raster must be derived using the d8_pointer tool and should have been extracted from a digital elevation model (DEM) that has been hydrologically pre-processed to remove topographic depressions and flat areas, e.g. using the breach_depressions_least_cost tool. By default, the flow pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools:

...
641281
3202
1684

If the pointer file contains ESRI flow direction values instead, the esri_pntr parameter must be specified.

The basins and watershed tools are similar in function but while the watershed tool identifies the upslope areas that drain to one or more user-specified outlet points, the basins tool automatically sets outlets to all grid cells situated along the edge of the data that do not have a defined flow direction (i.e. they do not have a lower neighbour). Notice that these edge outlets need not be situated along the edges of the flow-pointer raster, but rather along the edges of the region of valid data. That is, the DEM from which the flow-pointer has been extracted may incompletely fill the containing raster, if it is irregular shaped, and NoData regions may occupy the peripherals. Thus, the entire region of valid data in the flow pointer raster will be divided into a set of mutually exclusive basins using this tool.

See Also

watershed, d8_pointer, breach_depressions_least_cost

Function Signature

def basins(self, d8_pntr: Raster, esri_pntr: bool = False) -> Raster: ...

bilateral_filter

This tool can be used to perform an edge-preserving smoothing filter, or bilateral filter, on an image. A bilateral filter can be used to emphasize the longer-range variability in an image, effectively acting to smooth the image, while reducing the edge blurring effect common with other types of smoothing filters. As such, this filter is very useful for reducing the noise in an image. Bilateral filtering is a non-linear filtering technique introduced by Tomasi and Manduchi (1998). The algorithm operates by convolving a kernel of weights with each grid cell and its neighbours in an image. The bilateral filter is related to Gaussian smoothing, in that the weights of the convolution kernel are partly determined by the 2-dimensional Gaussian (i.e. normal) curve, which gives stronger weighting to cells nearer the kernel centre. Unlike the gaussian_filter, however, the bilateral kernel weightings are also affected by their similarity to the intensity value of the central pixel. Pixels that are very different in intensity from the central pixel are weighted less, also based on a Gaussian weight distribution. Therefore, this non-linear convolution filter is determined by the spatial and intensity domains of a localized pixel neighborhood.

The heavier weighting given to nearer and similar-valued pixels makes the bilateral filter an attractive alternative for image smoothing and noise reduction compared to the much-used Mean filter. The size of the filter is determined by setting the standard deviation distance parameter (sigma_dist); the larger the standard deviation the larger the resulting filter kernel. The standard deviation can be any number in the range 0.5-20 and is specified in the unit of pixels. The standard deviation intensity parameter (sigma_int), specified in the same units as the z-values, determines the intensity domain contribution to kernel weightings.

References

Tomasi, C., & Manduchi, R. (1998, January). Bilateral filtering for gray and color images. In null (p. 839). IEEE.

See Also

edge_preserving_mean_filter

Function Signature

def bilateral_filter(self, raster: Raster, sigma_dist: float = 0.75, sigma_int: float = 1.0) -> Raster: ...

block_maximum

Creates a raster grid based on a set of vector points and assigns grid values using a block maximum scheme.

Function Signature

def block_maximum(self, points: Vector, field_name: str = "FID", use_z: bool = False, cell_size: float = 0.0, base_raster: Raster = None) -> Raster: ...

block_minimum

Creates a raster grid based on a set of vector points and assigns grid values using a block minimum scheme.

Function Signature

def block_minimum(self, points: Vector, field_name: str = "FID", use_z: bool = False, cell_size: float = 0.0, base_raster: Raster = None) -> Raster: ...

bool_and

This tool is a Boolean AND operator, i.e. it works on True or False (1 and 0) values. Grid cells for which the first and second input rasters (input1; input2) have True values are assigned 1 in the output raster, otherwise grid cells are assigned a value of 0. All non-zero values in the input rasters are considered to be True, while all zero-valued grid cells are considered to be False. Grid cells containing NoData values in either of the input rasters will be assigned a NoData value in the output raster.

See Also

bool_not, bool_or, bool_xor

Function Signature

def bool_and(self, input1: Raster, input2: Raster) -> Raster: ...

bool_not

This tool is a Boolean NOT operator, i.e. it works on True or False (1 and 0) values. Grid cells for which the first input raster (input1) has a True value and the second raster (input2) has a False value are assigned 0 in the output raster, otherwise grid cells are assigned a value of 0. All non-zero values in the input rasters are considered to be True, while all zero-valued grid cells are considered to be False. Grid cells containing NoData values in either of the input rasters will be assigned a NoData value in the output raster. Notice that the Not operator is asymmetrical, and the order of inputs matters.

See Also

bool_and, bool_or, bool_xor

Function Signature

def bool_not(self, input1: Raster, input2: Raster) -> Raster: ...

bool_or

This tool is a Boolean OR operator, i.e. it works on True or False (1 and 0) values. Grid cells for which the either the first or second input rasters (input1; input2) have a True value are assigned 1 in the output raster, otherwise grid cells are assigned a value of 0. All non-zero values in the input rasters are considered to be True, while all zero-valued grid cells are considered to be False. Grid cells containing NoData values in either of the input rasters will be assigned a NoData value in the output raster.

See Also

bool_and, bool_not, bool_xor

Function Signature

def bool_or(self, input1: Raster, input2: Raster) -> Raster: ...

bool_xor

This tool is a Boolean XOR operator, i.e. it works on True or False (1 and 0) values. Grid cells for which either the first or second input rasters (input1; input2) have a True value but not both are assigned 1 in the output raster, otherwise grid cells are assigned a value of 0. All non-zero values in the input rasters are considered to be True, while all zero-valued grid cells are considered to be False. Grid cells containing NoData values in either of the input rasters will be assigned a NoData value in the output raster. Notice that the Not operator is asymmetrical, and the order of inputs matters.

See Also

bool_and, bool_not, bool_or

Function Signature

def bool_xor(self, input1: Raster, input2: Raster) -> Raster: ...

boundary_shape_complexity

This tools calculates a type of shape complexity index for raster objects, focused on the complexity of the boundary of polygons. The index uses the line_thinning tool to estimate a skeletonized network for each input raster polygon. The Boundary Shape Complexity (BSC) index is then calculated as the percentage of the skeletonized network belonging to exterior links. Polygons with more complex boundaries will possess more branching skeletonized networks, with each spur in the boundary possessing a short exterior branch. The two longest exterior links in the network are considered to be part of the main network. Therefore, polygons of complex shaped boundaries will have a higher percentage of their skeleton networks consisting of exterior links. It is expected that simple convex hulls should have relatively low BSC index values.

Objects in the input raster (input) are designated by their unique identifiers. Identifier values should be positive, non-zero whole numbers.

See Also

shape_complexity_index_raster, line_thinning

Function Signature

def boundary_shape_complexity(self, raster: Raster) -> Raster: ...

breach_depressions_least_cost

This tool can be used to perform a type of optimal depression breaching to prepare a digital elevation model (DEM) for hydrological analysis. Depression breaching is a common alternative to depression filling (fill_depressions) and often offers a lower-impact solution to the removal of topographic depressions. This tool implements a method that is loosely based on the algorithm described by Lindsay and Dhun (2015), furthering the earlier algorithm with efficiency optimizations and other significant enhancements. The approach uses a least-cost path analysis to identify the breach channel that connects pit cells (i.e. grid cells for which there is no lower neighbour) to some distant lower cell. Prior to breaching and in order to minimize the depth of breach channels, all pit cells are rised to the elevation of the lowest neighbour minus a small heigh value. Here, the cost of a breach path is determined by the amount of elevation lowering needed to cut the breach channel through the surrounding topography.

The user must specify the name of the input DEM file (dem), the output breached DEM file (output), the maximum search window radius (dist), the optional maximum breach cost (max_cost), and an optional flat height increment value (flat_increment). Notice that if the flat_increment parameter is not specified, the small number used to ensure flow across flats will be calculated automatically, which should be preferred in most applications of the tool. The tool operates by performing a least-cost path analysis for each pit cell, radiating outward until the operation identifies a potential breach destination cell or reaches the maximum breach length parameter. If a value is specified for the optional max_cost parameter, then least-cost breach paths that would require digging a channel that is more costly than this value will be left unbreached. The flat increment value is used to ensure that there is a monotonically descending path along breach channels to satisfy the necessary condition of a downslope gradient for flowpath modelling. It is best for this value to be a small value. If left unspecified, the tool with determine an appropriate value based on the range of elevation values in the input DEM, which should be the case in most applications. Notice that the need to specify these very small elevation increment values is one of the reasons why the output DEM will always be of a 64-bit floating-point data type, which will often double the storage requirements of a DEM (DEMs are often store with 32-bit precision). Lastly, the user may optionally choose to apply depression filling (fill) on any depressions that remain unresolved by the earlier depression breaching operation. This filling step uses an efficient filling method based on flooding depressions from their pit cells until outlets are identified and then raising the elevations of flooded cells back and away from the outlets.

The tool can be run in two modes, based on whether the min_dist is specified. If the min_dist flag is specified, the accumulated cost (accum2) of breaching from cell1 to cell2 along a channel issuing from pit is calculated using the traditional cost-distance function:

cost1 = z1 - (zpit + l × s)

cost2 = z2 - [zpit + (l + 1)s]

accum2 = accum1 + g(cost1 + cost2) / 2.0

where cost1 and cost2 are the costs associated with moving through cell1 and cell2 respectively, z1 and z2 are the elevations of the two cells, zpit is the elevation of the pit cell, l is the length of the breach channel to cell1, g is the grid cell distance between cells (accounting for diagonal distances), and s is the small number used to ensure flow across flats. If the min_dist flag is not present, the accumulated cost is calculated as:

accum2 = accum1 + cost2

That is, without the min_dist flag, the tool works to minimize elevation changes to the DEM caused by breaching, without considering the distance of breach channels. Notice that the value max_cost, if specified, should account for this difference in the way cost/cost-distances are calculated. The first cell in the least-cost accumulation operation that is identified for which cost2 <= 0.0 is the target cell to which the breach channel will connect the pit along the least-cost path.

In comparison with the breach_depressions_least_cost tool, this breaching method often provides a more satisfactory, lower impact, breaching solution and is often more efficient. It is therefore advisable that users try the breach_depressions_least_cost tool to remove depressions from their DEMs first. This tool is particularly well suited to breaching through road embankments. There are instances when a breaching solution is inappropriate, e.g. when a very deep depression such as an open-pit mine occurs in the DEM and long, deep breach paths are created. Often restricting breaching with the max_cost parameter, combined with subsequent depression filling (fill) can provide an adequate solution in these cases. Nonetheless, there are applications for which full depression filling using the fill_depressions tool may be preferred.

Reference

Lindsay J, Dhun K. 2015. Modelling surface drainage patterns in altered landscapes using LiDAR. International Journal of Geographical Information Science, 29: 1-15. DOI: 10.1080/13658816.2014.975715

See Also

breach_depressions_least_cost, fill_depressions, cost_pathway

Function Signature

def breach_depressions_least_cost(self, dem: Raster, max_cost: float = float('inf'), max_dist: int = 100, flat_increment: float = float('nan'), fill_deps: bool = False, minimize_dist: bool = False) -> Raster: ...

breach_single_cell_pits

This tool calculates the average slope gradient (i.e. slope steepness in degrees) of the flowpaths that pass through each grid cell in an input digital elevation model (DEM). The user must specify the name of a DEM raster (dem). It is important that this DEM is pre-processed to remove all topographic depressions and flat areas using a tool such as breach_depressions_least_cost. Several intermediate rasters are created and stored in memory during the operation of this tool, which may limit the size of DEM that can be processed, depending on available system resources.

See Also

average_upslope_flowpath_length, breach_depressions_least_cost

Function Signature

def breach_single_cell_pits(self, dem: Raster) -> Raster: ...

buffer_raster

This tool can be used to identify an area of interest within a specified distance of features of interest in a raster data set.

The Euclidean distance (i.e. straight-line distance) is calculated between each grid cell and the nearest 'target cell' in the input image. Distance is calculated using the efficient method of Shih and Wu (2004). Target cells are all non-zero, non-NoData grid cells. Because NoData values in the input image are assigned the NoData value in the output image, the only valid background value in the input image is zero.

The user must specify the input and output image names, the desired buffer size (size), and, optionally, whether the distance units are measured in grid cells (i.e. gridcells flag). If the gridcells flag is not specified, the linear units of the raster's coordinate reference system will be used.

Reference

Shih FY and Wu Y-T (2004), Fast Euclidean distance transformation in two scans using a 3 x 3 neighborhood, Computer Vision and Image Understanding, 93: 195-205.

See Also

euclidean_distance

Function Signature

def buffer_raster(self, input: Raster, buffer_size: float, grid_cells_units: bool = False) -> Raster: ...

burn_streams_at_roads

This tool decrements (lowers) the elevations of pixels within an input digital elevation model (DEM) (dem) along an input vector stream network (streams) at the sites of road (roads) intersections. In addition to the input data layers, the user must specify the output raster DEM (output), and the maximum road embankment width (width), in map units. The road width parameter is used to determine the length of channel along stream lines, at the junctions between streams and roads, that the burning (i.e. decrementing) operation occurs. The algorithm works by identifying stream-road intersection cells, then traversing along the rasterized stream path in the upstream and downstream directions by half the maximum road embankment width. The minimum elevation in each stream traversal is identified and then elevations that are higher than this value are lowered to the minimum elevation during a second stream traversal.

Reference

Lindsay JB. 2016. The practice of DEM stream burning revisited. Earth Surface Processes and Landforms, 41(5): 658–668. DOI: 10.1002/esp.3888

See Also

raster_streams_to_vector, rasterize_streams

Function Signature

def burn_streams_at_roads(self, dem: Raster, streams: Vector, roads: Vector, road_width: float) -> Raster: ...

centroid_raster

This tool calculates the centroid, or average location, of raster polygon objects. For vector features, use the centroid_vector tool instead.

See Also

centroid_vector

Function Signature

def centroid_raster(self, input: Raster) -> Tuple[Raster, str]: ...

centroid_vector

This can be used to identify the centroid point of a vector polyline or polygon feature or a group of vector points. The output is a vector shapefile of points. For multi-part polyline or polygon features, the user can optionally specify whether to identify the centroid of each part. The default is to treat multi-part features a single entity.

For raster features, use the Centroid tool instead.

See Also

Centroid, medoid

Function Signature

def centroid_vector(self, input: Vector) -> Vector: ...

change_vector_analysis

Change Vector Analysis (CVA) is a change detection method that characterizes the magnitude and change direction in spectral space between two times. A change vector is the difference vector between two vectors in n-dimensional feature space defined for two observations of the same geographical location (i.e. corresponding pixels) during two dates. The CVA inputs include the set of raster images corresponding to the multispectral data for each date. Note that there must be the same number of image files (bands) for the two dates and they must be entered in the same order, i.e. if three bands, red, green, and blue are entered for date one, these same bands must be entered in the same order for date two.

CVA outputs two image files. The first image contains the change vector length, i.e. magnitude, for each pixel in the multi-spectral dataset. The second image contains information about the direction of the change event in spectral feature space, which is related to the type of change event, e.g. deforestation will likely have a different change direction than say crop growth. The vector magnitude is a continuous numerical variable. The change vector direction is presented in the form of a code, referring to the multi-dimensional sector in which the change vector occurs. A text output will be produced to provide a key describing sector codes, relating the change vector to positive or negative shifts in n-dimensional feature space.

It is common to apply a simple thresholding operation on the magnitude data to determine 'actual' change (i.e. change above some assumed level of error). The type of change (qualitatively) is then defined according to the corresponding sector code. Jensen (2015) provides a useful description of this approach to change detection.

Reference

Jensen, J. R. (2015). Introductory Digital Image Processing: A Remote Sensing Perspective.

See Also

write_function_memory_insertion

Function Signature

def change_vector_analysis(self, date1_rasters: List[Raster], date2_rasters: List[Raster]) -> Tuple[Raster, Raster, str]: ...

circular_variance_of_aspect

This tool can be used to calculate the circular variance (i.e. one minus the mean resultant length) of aspect for a digital elevation model (DEM). This is a measure of how variable slope aspect is within a local neighbourhood of a specified size (filter). circular_variance_of_aspect is therefore a measure of surface shape complexity, or texture. It will take a value of 0.0 for smooth sites and near 1.0 in areas of high surface roughness or complex topography.

The local neighbourhood size (filter) must be any odd integer equal to or greater than three. Grohmann et al. (2010) found that vector dispersion, a related measure of angular variance, increases monotonically with scale. This is the result of the angular dispersion measure integrating (accumulating) all of the surface variance of smaller scales up to the test scale. A more interesting scale relation can therefore be estimated by isolating the amount of surface complexity associated with specific scale ranges. That is, at large spatial scales, the metric should reflect the texture of large-scale landforms rather than the accumulated complexity at all smaller scales, including microtopographic roughness. As such, this tool normalizes the surface complexity of scales that are smaller than the filter size by applying Gaussian blur (with a standard deviation of one-third the filter size) to the DEM prior to calculating circular_variance_of_aspect. In this way, the resulting distribution is able to isolate and highlight the surface shape complexity associated with landscape features of a similar scale to that of the filter size.

This tool makes extensive use of integral images (i.e. summed-area tables) and parallel processing to ensure computational efficiency. It may, however, require substantial memory resources when applied to larger DEMs.

References

Grohmann, C. H., Smith, M. J., & Riccomini, C. (2010). Multiscale analysis of topographic surface roughness in the Midland Valley, Scotland. IEEE Transactions on Geoscience and Remote Sensing, 49(4), 1200-1213.

See Also

aspect, spherical_std_dev_of_normals, multiscale_roughness, edge_density, surface_area_ratio, ruggedness_index

Function Signature

def circular_variance_of_aspect(self, dem: Raster, filter_size: int = 11) -> Raster: ...

classify_buildings_in_lidar

This tool can be used to assign the building class (classification value 6) to all points within an input LiDAR point cloud (input) that are contained within the polygons of an input buildings footprint vector (buildings). The tool performs a simple point-in-polygon operation to determine membership. The two inputs (i.e. the LAS file and vector) must share the same map projection. Furthermore, any error in the definition of the building footprints will result in misclassified points in the output LAS file (output). In particular, if the footprints extend slightly beyond the actual building, ground points situated adjacent to the building will be incorrectly classified. Thus, care must be taken in digitizing building footprint polygons. Furthermore, where there are tall trees that overlap significantly with the building footprint, these vegetation points will also be incorrectly assigned the building class value.

See Also

filter_lidar_classes, lidar_ground_point_filter, clip_lidar_to_polygon

Function Signature

def classify_buildings_in_lidar(self, in_lidar: Lidar, building_footprints: Vector) -> Lidar: ...

classify_overlap_points

This tool can be used to flag points within an input LiDAR file (input) that overlap with other nearby points from different flightlines, i.e. to identify overlap points. The flightline associated with a LiDAR point is assumed to be contained within the point's Point Source ID (PSID) property. If the PSID property is not set, or has been lost, users may with to apply the RecoverFlightlineInfo tool prior to running flightline_overlap.

Areas of multiple flightline overlap tend to have point densities that are far greater than areas of single flightlines. This can produce suboptimal results for applications that assume regular point distribution, e.g. in point classification operations.

The tool works by applying a square grid over the extent of the input LiDAR file. The grid cell size is determined by the user-defined resolution parameter. Grid cells containing multiple PSIDs, i.e. with more than one flightline, are then identified. Overlap points within these grid cells can then be flagged on the basis of a user-defined criterion. The flagging options include the following:

CriterionOverlap Point Definition
max scan angleAll points that share the PSID of the point with the maximum absolute scan angle
not min point source IDAll points with a different PSID to that of the point with the lowest PSID
not min timeAll points with a different PSID to that of the point with the minimum GPS time
multiple point source IDsAll points in grid cells with multiple PSIDs, i.e. all overlap points.

Note that the max scan angle criterion may not be appropriate when more than two flightlines overlap, since it will result in only flagging points from one of the multiple flightlines.

It is important to set the resolution parameter appropriately, as setting this value too high will yield the filtering of points in non-overlap areas, and setting the resolution to low will result in fewer than expected points being flagged. An appropriate resolution size value may require experimentation, however a value that is 2-3 times the nominal point spacing has been previously recommended. The nominal point spacing can be determined using the lidar_info tool.

By default, all flagged overlap points are reclassified in the output LiDAR file (output) to class 12. Alternatively, if the user specifies the filter parameter, then each overlap point will be excluded from the output file. Classified overlap points may also be filtered from LiDAR point clouds using the FilterLidar tool.

Note that this tool is intended to be applied to LiDAR tile data containing points that have been merged from multiple overlapping flightlines. It is commonly the case that airborne LiDAR data from each of the flightlines from a survey are merged and then tiled into 1 km2 tiles, which are the target dataset for this tool.

See Also

flightline_overlap, RecoverFlightlineInfo, FilterLidar, lidar_info

Function Signature

def classify_overlap_points(self, in_lidar: Lidar, resolution: float = 1.0, overlap_criterion: str = "max scan angle", filter: bool = False) -> Lidar: ...

clean_vector

Description

This tool can be used to remove all features in Shapefiles that are of the null VectorGeometryType. It also removes line features with fewer than two vertices and polygon features with fewer than three vertices.

Parameters

input (Vector): The input Vector object

Returns

Vector: the returning value

Function Signature

def clean_vector(self, input: Vector) -> Vector: ...

clip

This tool will extract all the features, or parts of features, that overlap with the features of the clip vector file. The clipping operation is one of the most common vector overlay operations in GIS and effectively imposes the boundary of the clip layer on a set of input vector features, or target features. The operation is sometimes likened to a 'cookie-cutter'. The input vector file can be of any feature type (i.e. points, lines, polygons), however, the clip vector must consist of polygons.

See Also

erase

Function Signature

def clip(self, input: Vector, clip_layer: Vector) -> Vector: ...

clip_lidar_to_polygon

This tool can be used to isolate, or clip, all of the LiDAR points in a LAS file (input) contained within one or more vector polygon features. The user must specify the name of the input clip file (--polygons), which must be a vector of a Polygon base shape type. The clip file may contain multiple polygon features and polygon hole parts will be respected during clipping, i.e. LiDAR points within polygon holes will be removed from the output LAS file.

Use the erase_polygon_from_lidar tool to perform the complementary operation of removing points from a LAS file that are contained within a set of polygons.

See Also

erase_polygon_from_lidar, FilterLidar, clip, clip_raster_to_polygon

Function Signature

def clip_lidar_to_polygon(self, input: Lidar, polygons: Vector) -> Lidar: ...

clip_raster_to_polygon

This tool can be used to clip an input raster (input) to the extent of a vector polygon (shapefile). The user must specify the name of the input clip file (polygons), which must be a vector of a Polygon base shape type. The clip file may contain multiple polygon features. Polygon hole parts will be respected during clipping, i.e. polygon holes will be removed from the output raster by setting them to a NoData background value. Raster grid cells that fall outside of a polygons in the clip file will be assigned the NoData background value in the output file. By default, the output raster will be cropped to the spatial extent of the clip file, unless the maintain_dimensions parameter is used, in which case the output grid extent will match that of the input raster. The grid resolution of output raster is the same as the input raster.

It is very important that the input raster and the input vector polygon file share the same projection. The result is unlikely to be satisfactory otherwise.

See Also

erase_polygon_from_raster

Function Signature

def clip_raster_to_polygon(self, raster: Raster, polygons: Vector, maintain_dimensions: bool = False) -> Raster: ...

closing

This tool performs a closing operation on an input greyscale image (input). A closing is a mathematical morphology operation involving an erosion (minimum filter) of a dilation (maximum filter) set. closing operations, together with the opening operation, is frequently used in the fields of computer vision and digital image processing for image noise removal. The user must specify the size of the moving window in both the x and y directions (filterx and filtery).

See Also

opening, tophat_transform

Function Signature

def closing(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

clump

This tool re-categorizes data in a raster image by grouping cells that form discrete, contiguous areas into unique categories. Essentially this will produce a patch map from an input categorical raster, assigning each feature unique identifiers. The input raster should either be Boolean (1's and 0's) or categorical. The input raster could be created using the reclass tool or one of the comparison operators (GreaterThan, LessThan, EqualTo, NotEqualTo). Use the treat zeros as background cells options (zero_back) if you would like to only assigned contiguous groups of non-zero values in the raster unique identifiers. Additionally, inter-cell connectivity can optionally include diagonally neighbouring cells if the diag flag is specified.

See Also

reclass, GreaterThan, LessThan, EqualTo, NotEqualTo

Function Signature

def clump(self, raster: Raster, diag: bool = False, zero_background: bool = False) -> Raster: ...

compactness_ratio

The compactness ratio is an indicator of polygon shape complexity. The compactness ratio is defined as the polygon area divided by its perimeter. Unlike some other shape parameters (e.g. ShapeComplexityIndex), compactness ratio does not standardize to a simple Euclidean shape. Although widely used for landscape analysis, compactness ratio, like its inverse, the perimeter_area_ratio, exhibits the undesirable property of polygon size dependence (Mcgarigal et al. 2002). That is, holding shape constant, an increase in polygon size will cause a change in the compactness ratio.

The output data will be contained in the input vector's attribute table as a new field (COMPACT).

See Also

perimeter_area_ratio, ShapeComplexityIndex, related_circumscribing_circle

Function Signature

def compactness_ratio(self, input: Vector) -> Vector: ...

conservative_smoothing_filter

This tool performs a conservative smoothing filter on a raster image. A conservative smoothing filter can be used to remove short-range variability in an image, effectively acting to smooth the image. It is particularly useful for eliminating local spikes and reducing the noise in an image. The algorithm operates by calculating the minimum and maximum neighbouring values surrounding a grid cell. If the cell at the centre of the kernel is greater than the calculated maximum value, it is replaced with the maximum value in the output image. Similarly, if the cell value at the kernel centre is less than the neighbouring minimum value, the corresponding grid cell in the output image is replaced with the minimum value. This filter tends to alter an image very little compared with other smoothing filters such as the mean_filter, edge_preserving_mean_filter, bilateral_filter, median_filter, gaussian_filter, or olympic_filter.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

See Also

mean_filter, edge_preserving_mean_filter, bilateral_filter, median_filter, gaussian_filter, olympic_filter

Function Signature

def conservative_smoothing_filter(self, raster: Raster, filter_size_x: int = 3, filter_size_y: int = 3) -> Raster: ...

construct_vector_tin

This tool creates a vector triangular irregular network (TIN) for a set of vector points (input) using a 2D Delaunay triangulation algorithm. TIN vertex heights can be assigned based on either a field in the vector's attribute table (field), or alternatively, if the vector is of a z-dimension VectorGeometryTypeDimension, the point z-values may be used for vertex heights (use_z). For LiDAR points, use the lidar_construct_vector_tin tool instead.

Triangulation often creates very long, narrow triangles near the edges of the data coverage, particularly in convex regions along the data boundary. To avoid these spurious triangles, the user may optionally specify the maximum allowable edge length of a triangular facet (max_triangle_edge_length).

See Also

lidar_construct_vector_tin

Function Signature

def construct_vector_tin(self, input_points: Vector, field_name: str = "FID", use_z: bool = False, max_triangle_edge_length: float = float('inf')) -> Vector: ...

contours_from_points

This tool creates a contour coverage from a set of input points (input). The user must specify the contour interval (interval) and optionally, the base contour value (base). The degree to which contours are smoothed is controlled by the Smoothing Filter Size parameter (smooth). This value, which determines the size of a mean filter applied to the x-y position of vertices in each contour, should be an odd integer value, e.g. 3, 5, 7, 9, 11, etc. Larger values will result in smoother contour lines.

See Also

contours_from_raster

Function Signature

def contours_from_points(self, input: Vector, field_name: str = "", use_z_values: bool = False, max_triangle_edge_length: float = float('inf'), contour_interval: float = 10.0, base_contour: float = 0.0, smoothing_filter_size: int = 9) -> Vector: ...

contours_from_raster

This tool can be used to create a vector contour coverage from an input raster surface model (input), such as a digital elevation model (DEM). The user must specify the contour interval (interval) and optionally, the base contour value (base). The degree to which contours are smoothed is controlled by the Smoothing Filter Size parameter (smooth). This value, which determines the size of a mean filter applied to the x-y position of vertices in each contour, should be an odd integer value, e.g. 3, 5, 7, 9, 11, etc. Larger values will result in smoother contour lines. The tolerance parameter (tolerance) controls the amount of line generalization. That is, vertices in a contour line will be selectively removed from the line if they do not result in an angular deflection in the line's path of at least this threshold value. Increasing this value can significantly decrease the size of the output contour vector file, at the cost of generating straighter contour line segments.

See Also

raster_to_vector_polygons

Function Signature

def contours_from_raster(self, raster_surface: Raster, contour_interval: float = 10.0, base_contour: float = 0.0, smoothing_filter_size: int = 9, deflection_tolerance: float = 10.0) -> Vector: ...

convert_nodata_to_zero

Description

This tool can be used to change the value within the grid cells of a raster (input) that contain NoData to zero. The most common reason for using this tool is to change the background region of a raster image such that it can be included in analysis since NoData values are usually ignored by by most tools. This change, however, will result in the background no longer displaying transparently in most GIS. This change can be reversed using the set_nodata_value tool.

See Also

set_nodata_value, Raster.is_nodata

Parameters

raster (Raster): The input Raster object

Returns

Raster: the returning value

Function Signature

def convert_nodata_to_zero(self, raster: Raster) -> Raster: ...

corner_detection

This tool identifies corner patterns in boolean images using hit-and-miss pattern matching. Foreground pixels in the input image (input) are designated by any positive, non-zero values. Zero-valued and NoData-valued grid cells are interpreted by the algorithm as background values.

Reference

Fisher, R, Brown, N, Cammas, N, Fitzgibbon, A, Horne, S, Koryllos, K, Murdoch, A, Robertson, J, Sharman, T, Strachan, C, 2004. Hypertext Image Processing Resource. online: http://homepages.inf.ed.ac.uk/rbf/HIPR2/hitmiss.htm

Function Signature

def corner_detection(self, raster: Raster) -> Raster: ...

correct_vignetting

This tool can be used to reduce vignetting within an image. Vignetting refers to the reduction of image brightness away from the image centre (i.e. the principal point). Vignetting is a radiometric distortion resulting from lens characteristics. The algorithm calculates the brightness value in the output image (BVout) as:

BVout = BVin / [cos^n(arctan(d / f))]

Where d is the photo-distance from the principal point in millimetres, f is the focal length of the camera, in millimeters, and n is a user-specified parameter. Pixel distances are converted to photo-distances (in millimetres) using the specified image width, i.e. distance between left and right edges (mm). For many cameras, 4.0 is an appropriate value of the n parameter. A second pass of the image is used to rescale the output image so that it possesses the same minimum and maximum values as the input image.

If an RGB image is input, the analysis will be performed on the intensity component of the HSI transform.

Function Signature

def correct_vignetting(self, image: Raster, principal_point: Vector, focal_length: float = 304.8, image_width: float = 228.6, n_param: float = 4.0) -> Raster: ...

cost_allocation

This tool can be used to identify the 'catchment area' of each source grid cell in a cost-distance analysis. The user must specify the names of the input source and back-link raster files. Source cells (i.e. starting points for the cost-distance or least-cost path analysis) are designated as all positive, non-zero valued grid cells in the source raster. A back-link raster file can be created using the cost_distance tool and is conceptually similar to the D8 flow-direction pointer raster grid in that it describes the connectivity between neighbouring cells on the accumulated cost surface.

NoData values in the input back-link image are assigned NoData values in the output image.

See Also

cost_distance, cost_pathway, euclidean_allocation

Function Signature

def cost_allocation(self, source: Raster, backlink: Raster) -> Raster: ...

cost_distance

This tool can be used to perform cost-distance or least-cost pathway analyses. Specifically, this tool can be used to calculate the accumulated cost of traveling from the 'source grid cell' to each other grid cell in a raster dataset. It is based on the costs associated with traveling through each cell along a pathway represented in a cost (or friction) surface. If there are multiple source grid cells, each cell in the resulting cost-accumulation surface will reflect the accumulated cost to the source cell that is connected by the minimum accumulated cost-path. The user must specify the names of the raster file containing the source cells (source), the raster file containing the cost surface information (cost), the output cost-accumulation surface raster (out_accum), and the output back-link raster (out_backlink). Source cells are designated as all positive, non-zero valued grid cells in the source raster. The cost (friction) raster can be created by combining the various cost factors associated with the specific problem (e.g. slope gradient, visibility, etc.) using a raster calculator or the weighted_overlay tool.

While the cost-accumulation surface raster can be helpful for visualizing the three-dimensional characteristics of the 'cost landscape', it is actually the back-link raster that is used as inputs to the other two cost-distance tools, cost_allocation and cost_pathway, to determine the least-cost linkages among neighbouring grid cells on the cost surface. If the accumulated cost surface is analogous to a digital elevation model (DEM) then the back-link raster is equivalent to the D8 flow-direction pointer. In fact, it is created in a similar way and uses the same convention for designating 'flow directions' between neighbouring grid cells. The algorithm for the cost distance accumulation operation uses a type of priority-flood method similar to what is used for depression filling and flow accumulation operations.

NoData values in the input cost surface image are ignored during processing and assigned NoData values in the outputs. The output cost accumulation raster is of the float data type and continuous data scale.

See Also

cost_allocation, cost_pathway, weighted_overlay

Function Signature

def cost_distance(self, source: Raster, cost: Raster) -> Tuple[Raster, Raster]: ...

cost_pathway

This tool can be used to map the least-cost pathway connecting each destination grid cell in a cost-distance analysis to a source cell. The user must specify the names of the input destination and back-link raster files. Destination cells (i.e. end points for the least-cost path analysis) are designated as all positive, non-zero valued grid cells in the destination raster. A back-link raster file can be created using the cost_distance tool and is conceptually similar to the D8 flow-direction pointer raster grid in that it describes the connectivity between neighbouring cells on the accumulated cost surface. All background grid cells in the output image are assigned the NoData value.

NoData values in the input back-link image are assigned NoData values in the output image.

See Also

cost_distance, cost_allocation

Function Signature

def cost_pathway(self, destination: Raster, backlink: Raster, zero_background: bool = False) -> Raster: ...

count_if

This tool counts the number of occurrences of a specified value (value) in a stack of input rasters (inputs). Each grid cell in the output raster (output) will contain the number of occurrences of the specified value in the stack of corresponding cells in the input image. At least two input rasters are required to run this tool. Each of the input rasters must share the same number of rows and columns and spatial extent. An error will be issued if this is not the case.

See Also

pick_from_list

Function Signature

def count_if(self, input_rasters: List[Raster], comparison_value: float) -> Raster: ...

create_colour_composite

This tool can be used to create a colour-composite image from three bands of multi-spectral imagery. The user must input images to enter into the red, green, and blue channels of the resulting composite image. The output image uses the 32-bit aRGB colour model, and therefore, in addition to red, green and blue bands, the user may optionally specify a fourth image that will be used to determine pixel opacity (the 'a' channel). If no opacity image is specified, each pixel will be opaque. This can be useful for cropping an image to an irregular-shaped boundary. The opacity channel can also be used to create transparent gradients in the composite image.

A balance contrast enhancement (BCE) can optionally be performed on the bands prior to creation of the colour composite. While this operation will add to the runtime of create_colour_composite, if the individual input bands have not already had contrast enhancements, then it is advisable that the BCE option be used to improve the quality of the resulting colour composite image.

NoData values in any of the input images are assigned NoData values in the output image and are not taken into account when performing the BCE operation. Please note, not all images have NoData values identified. When this is the case, and when the background value is 0 (often the case with multispectral imagery), then the create_colour_composite tool can be told to ignore zero values using the zeros flag.

See Also

balance_contrast_enhancement, split_colour_composite

Function Signature

def create_colour_composite(self, red: Raster, green: Raster, blue: Raster, opacity: Raster = None, enhance: bool = True, treat_zeros_as_nodata: bool = False) -> Raster: ...

create_plane

This tool can be used to create a new raster with values that are determined by the equation of a simple plane. The user must specify the name of a base raster (base) from which the output raster coordinate and dimensional information will be taken. In addition the user must specify the values of the planar slope gradient (S; gradient; aspect) in degrees, the planar slope direction or aspect (A; 0 to 360 degrees), and an constant value (k; constant). The equation of the plane is as follows:

Z = tan(S) × sin(A - 180) × X + tan(S) × cos(A - 180) × Y + k

where X and Y are the X and Y coordinates of each grid cell in the grid. Notice that A is the direction, or azimuth, that the plane is facing

Function Signature

def create_plane(self, base_file: Raster, gradient: float, aspect: float, constant: float) -> Raster: ...

crispness_index

The Crispness Index (C) provides a means of quantifying the crispness, or fuzziness, of a membership probability (MP) image. MP images describe the probability of each grid cell belonging to some feature or class. MP images contain values ranging from 0 to 1.

The index, as described by Lindsay (2006), is the ratio between the sum of the squared differences (from the image mean) in the MP image divided by the sum of the squared differences for the Boolean case in which the total probability, summed for the image, is arranged crisply.

C is closely related to a family of relative variation coefficients that measure variation in an MP image relative to the maximum possible variation (i.e. when the total probability is arranged such that grid cells contain only 1s or 0s). Notice that 0 < C < 1 and a low C-value indicates a nearly uniform spatial distribution of any probability value, and C = 1 indicates a crisp spatial probability distribution, containing only 1's and 0's.

C is calculated as follows:

C = SS_mp ∕ SS_B = [∑(pij − p-bar)^2] ∕ [ ∑pij(1 − p-bar)^2 + p2(RC − ∑pij)]

Note that there is an error in the original published equation. Specifically, the denominator read:

∑pij(1 - p_bar)^2 + p_bar^2 (RC - ∑pij)

instead of the original:

∑pij(1 - p_bar^2) - p_bar^2 (RC - ∑pij)

References

Lindsay, J. B. (2006). Sensitivity of channel mapping techniques to uncertainty in digital elevation data. International Journal of Geographical Information Science, 20(6), 669-692.

Function Signature

def crispness_index(self, raster: Raster, output_html_file: str) -> None: ...

cross_tabulation

This tool can be used to perform a cross-tabulation on two input raster images (i1 and i2) containing categorical data, i.e. classes. It will output a contingency table in HTML format (output). A contingency table, also known as a cross tabulation or crosstab, is a type of table that displays the multivariate frequency distribution of the variables. These tables provide a basic picture of the interrelation between two categorical variables and can help find interactions between them. cross_tabulation can provide useful information about the nature of land-use/land-cover (LULC) changes between two dates of classified multi-spectral satellite imagery. For example, the extent of urban expansion could be described using the information about the extent of pixels in an 'urban' class in Date 2 that were previously assigned to other classes (e.g. agricultural LULC categories) in the Date 1 imagery.

Both input images must share the same grid, as the analysis requires a comparison of a pair of images on a cell-by-cell basis. If a grid cell contains a NoData value in either of the input images, the cell will be excluded from the analysis.

Function Signature

def cross_tabulation(self, raster1: Raster, raster2: Raster, output_html_file: str) -> None: ...

csv_points_to_vector

This tool can be used to import a series of points contained within a comma-separated values (*.csv) file (input_file) into a vector shapefile of a POINT VectorGeometryType. The input file must be an ASCII text file with a .csv extensions. The tool will automatically detect the field data type; for numeric fields, it will also determine the appropriate length and precision. The user must specify the x-coordinate (x_field_num) and y-coordiante (y_field_num) fields. All fields are imported as attributes in the output (output) vector file. The tool assumes that the first line of the file is a header line from which field names are retrieved.

See Also

merge_table_with_csv, export_table_to_csv

Function Signature

def csv_points_to_vector(self, input_file: str, x_field_num: int = 0, y_field_num: int = 1, epsg: int = 0) -> Vector: ...

cumulative_distribution

This tool converts the values in an input image (input) into a cumulative distribution function. Therefore, the output raster (output) will contain the cumulative probability value (0-1) of of values equal to or less than the value in the corresponding grid cell in the input image. NoData values in the input image are not considered during the transformation and remain NoData values in the output image.

See Also

z_scores

Function Signature

def cumulative_distribution(self, raster: Raster) -> Raster: ...

d8_flow_accum

This tool is used to generate a flow accumulation grid (i.e. catchment area) using the D8 (O'Callaghan and Mark, 1984) algorithm. This algorithm is an example of single-flow-direction (SFD) method because the flow entering each grid cell is routed to only one downslope neighbour, i.e. flow divergence is not permitted. The user must specify the name of the input digital elevation model (DEM) or flow pointer raster (input) derived using the D8 or Rho8 method (d8_pointer, rho8_pointer). If an input DEM is used, it must have been hydrologically corrected to remove all spurious depressions and flat areas. DEM pre-processing is usually achieved using the breach_depressions_least_cost or fill_depressions tools. If a D8 pointer raster is input, the user must also specify the optional pntr flag. If the D8 pointer follows the Esri pointer scheme, rather than the default WhiteboxTools scheme, the user must also specify the optional esri_pntr flag.

In addition to the input DEM/pointer, the user must specify the output type. The output flow-accumulation can be 1) cells (i.e. the number of inflowing grid cells), catchment area (i.e. the upslope area), or specific contributing area (i.e. the catchment area divided by the flow width. The default value is cells. The user must also specify whether the output flow-accumulation grid should be log-tranformed (log), i.e. the output, if this option is selected, will be the natural-logarithm of the accumulated flow value. This is a transformation that is often performed to better visualize the contributing area distribution. Because contributing areas tend to be very high along valley bottoms and relatively low on hillslopes, when a flow-accumulation image is displayed, the distribution of values on hillslopes tends to be 'washed out' because the palette is stretched out to represent the highest values. Log-transformation provides a means of compensating for this phenomenon. Importantly, however, log-transformed flow-accumulation grids must not be used to estimate other secondary terrain indices, such as the wetness index, or relative stream power index.

Grid cells possessing the NoData value in the input DEM/pointer raster are assigned the NoData value in the output flow-accumulation image.

Reference

O'Callaghan, J. F., & Mark, D. M. 1984. The extraction of drainage networks from digital elevation data. Computer Vision, Graphics, and Image Processing, 28(3), 323-344.

See Also:

FD8FlowAccumulation, quinn_flow_accumulation, qin_flow_accumulation, DInfFlowAccumulation, MDInfFlowAccumulation, rho8_pointer, d8_pointer, breach_depressions_least_cost, fill_depressions

Function Signature

def d8_flow_accum(self, raster: Raster, out_type: str = "sca", log_transform: bool = False, clip: bool = False, input_is_pointer: bool = False, esri_pntr: bool = False) -> Raster: ...

d8_mass_flux

This tool can be used to perform a mass flux calculation using DEM-based surface flow-routing techniques. For example, it could be used to model the distribution of sediment or phosphorous within a catchment. Flow-routing is based on a D8 flow pointer (i.e. flow direction) derived from an input depresionless DEM (dem). The user must also specify the names of loading (loading), efficiency (efficiency), and absorption (absorption) rasters, as well as the output raster. Mass Flux operates very much like a flow-accumulation operation except that rather than accumulating catchment areas the algorithm routes a quantity of mass, the spatial distribution of which is specified within the loading image. The efficiency and absorption rasters represent spatial distributions of losses to the accumulation process, the difference being that the efficiency raster is a proportional loss (e.g. only 50% of material within a particular grid cell will be directed downslope) and the absorption raster is an loss specified as a quantity in the same units as the loading image. The efficiency image can range from 0 to 1, or alternatively, can be expressed as a percentage. The equation for determining the mass sent from one grid cell to a neighbouring grid cell is:

Outflowing Mass = (Loading - Absorption + Inflowing Mass) × Efficiency

This tool assumes that each of the three input rasters have the same number of rows and columns and that any NoData cells present are the same among each of the inputs.

See Also

DInfMassFlux

Function Signature

def d8_mass_flux(self, dem: Raster, loading: Raster, efficiency: Raster, absorption: Raster) -> Raster: ...

d8_pointer

This tool is used to generate a flow pointer grid using the simple D8 (O'Callaghan and Mark, 1984) algorithm. The user must specify the name (dem) of a digital elevation model (DEM) that has been hydrologically corrected to remove all spurious depressions and flat areas. DEM pre-processing is usually achieved using either the breach_depressions_least_cost or fill_depressions tool. The local drainage direction raster output (output) by this tool serves as a necessary input for several other spatial hydrology and stream network analysis tools in the toolset. Some tools will calculate this flow pointer raster directly from the input DEM.

By default, D8 flow pointers use the following clockwise, base-2 numeric index convention:

...
641281
3202
1684

Notice that grid cells that have no lower neighbours are assigned a flow direction of zero. In a DEM that has been pre-processed to remove all depressions and flat areas, this condition will only occur along the edges of the grid. If the pointer file contains ESRI flow direction values instead, the esri_pntr parameter must be specified.

Grid cells possessing the NoData value in the input DEM are assigned the NoData value in the output image.

Memory Usage

The peak memory usage of this tool is approximately 10 bytes per grid cell.

Reference

O'Callaghan, J. F., & Mark, D. M. (1984). The extraction of drainage networks from digital elevation data. Computer vision, graphics, and image processing, 28(3), 323-344.

See Also

DInfPointer, fd8_pointer, breach_depressions_least_cost, fill_depressions

Function Signature

def d8_pointer(self, dem: Raster, esri_pointer: bool = False) -> Raster: ...

depth_in_sink

This tool measures the depth that each grid cell in an input (dem) raster digital elevation model (DEM) lies within a sink feature, i.e. a closed topographic depression. A sink, or depression, is a bowl-like landscape feature, which is characterized by interior drainage and groundwater recharge. The depth_in_sink tool operates by differencing a filled DEM, using the same depression filling method as fill_depressions, and the original surface model.

In addition to the names of the input DEM (dem) and the output raster (output), the user must specify whether the background value (i.e. the value assigned to grid cells that are not contained within sinks) should be set to 0.0 (zero_background) Without this optional parameter specified, the tool will use the NoData value as the background value.

Reference

Antonić, O., Hatic, D., & Pernar, R. (2001). DEM-based depth in sink as an environmental estimator. Ecological Modelling, 138(1-3), 247-254.

See Also

fill_depressions

Function Signature

def depth_in_sink(self, dem: Raster, zero_background: bool = False) -> Raster: ...

deviation_from_mean_elevation

This tool can be used to calculate the difference between the elevation of each grid cell and the mean elevation of the centering local neighbourhood, normalized by standard deviation. Therefore, this index of topographic residual is essentially equivalent to a local z-score. This attribute measures the relative topographic position as a fraction of local relief, and so is normalized to the local surface roughness. DevFromMeanElev utilizes an integral image approach (Crow, 1984) to ensure highly efficient filtering that is invariant with filter size.

The user must input a digital elevation model (DEM) (dem) and the size of the neighbourhood in the x and y directions (filterx and filtery), measured in grid size.

While DeviationFromMeanElev calculates the deviation from mean elevation (DEV) at a single, user-defined scale, the max_elevation_deviation tool can be used to output the per-pixel maximum DEV value across a range of input scales.

See Also

DiffFromMeanElev, max_elevation_deviation

Function Signature

def deviation_from_mean_elevation(self, dem: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

diff_of_gaussians_filter

This tool can be used to perform a difference-of-Gaussians (DoG) filter on a raster image. In digital image processing, DoG is a feature enhancement algorithm that involves the subtraction of one blurred version of an image from another, less blurred version of the original. The blurred images are obtained by applying filters with Gaussian-weighted kernels of differing standard deviations to the input image (input). Blurring an image using a Gaussian-weighted kernel suppresses high-frequency spatial information and emphasizes lower-frequency variation. Subtracting one blurred image from the other preserves spatial information that lies between the range of frequencies that are preserved in the two blurred images. Thus, the difference-of-Gaussians is a band-pass filter that discards all but a specified range of spatial frequencies that are present in the original image.

The algorithm operates by differencing the results of convolving two kernels of weights with each grid cell and its neighbours in an image. The weights of the convolution kernels are determined by the 2-dimensional Gaussian (i.e. normal) curve, which gives stronger weighting to cells nearer the kernel centre. The size of the two convolution kernels are determined by setting the two standard deviation parameters (sigma1 and sigma2); the larger the standard deviation the larger the resulting filter kernel. The second standard deviation should be a larger value than the first, however if this is not the case, the tool will automatically swap the two parameters. Both standard deviations can range from 0.5-20.

The difference-of-Gaussians filter can be used to emphasize edges present in an image. Other edge-sharpening filters also operate by enhancing high-frequency detail, but because random noise also has a high spatial frequency, many of these sharpening filters tend to enhance noise, which can be an undesirable artifact. The difference-of-Gaussians filter can remove high-frequency noise while emphasizing edges. This filter can, however, reduce overall image contrast.

See Also

gaussian_filter, fast_almost_gaussian_filter, laplacian_filter, LaplacianOfGaussianFilter`

Function Signature

def diff_of_gaussians_filter(self, raster: Raster, sigma1: float = 2.0, sigma2: float = 4.0) -> Raster: ...

difference

This tool will remove all the overlapping features, or parts of overlapping features, between input and overlay vector files, outputting only the features that occur in one of the two inputs but not both. The Symmetrical Difference is related to the Boolean exclusive-or (XOR) operation in set theory and is one of the common vector overlay operations in GIS. The user must specify the names of the input and overlay vector files as well as the output vector file name. The tool operates on vector points, lines, or polygon, but both the input and overlay files must contain the same VectorGeometryType.

The Symmetrical Difference can also be derived using a combination of other vector overlay operations, as either (A union B) difference (A intersect B), or (A difference B) union (B difference A).

The attributes of the two input vectors will be merged in the output attribute table. Fields that are duplicated between the inputs will share a single attribute in the output. Fields that only exist in one of the two inputs will be populated by null in the output table. Multipoint VectorGeometryTypes however will simply contain a single output feature identifier (FID) attribute. Also, note that depending on the VectorGeometryType (polylines and polygons), Measure and Z ShapeDimension data will not be transferred to the output geometries. If the input attribute table contains fields that measure the geometric properties of their associated features (e.g. length or area), these fields will not be updated to reflect changes in geometry shape and size resulting from the overlay operation.

See Also

intersect, difference, union, clip, erase

Function Signature

def difference(self, input: Vector, overlay: Vector) -> Vector: ...

difference_from_mean_elevation

This tool can be used to calculate the difference between the elevation of each grid cell and the mean elevation of the centering local neighbourhood. This is similar to what a high-pass filter calculates for imagery data, but is intended to work with DEM data instead. This attribute measures the relative topographic position. DiffFromMeanElev utilizes an integral image approach (Crow, 1984) to ensure highly efficient filtering that is invariant with filter size.

The user must specify a digital elevation model (DEM) (dem) , and the size of the neighbourhood in the x and y directions (filterx and filtery), measured in grid size.

While DevFromMeanElev calculates the DIFF at a single, user-defined scale, the max_difference_from_mean tool can be used to output the per-pixel maximum DIFF value across a range of input scales.

See Also

DevFromMeanElev, max_difference_from_mean

Function Signature

def difference_from_mean_elevation(self, dem: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

dinf_flow_accum

This tool is used to generate a flow accumulation grid (i.e. contributing area) using the D-infinity algorithm (Tarboton, 1997). This algorithm is an examples of a multiple-flow-direction (MFD) method because the flow entering each grid cell is routed to one or two downslope neighbour, i.e. flow divergence is permitted. The user must specify the name of the input digital elevation model or D-infinity pointer raster (input). If an input DEM is specified, the DEM should have been hydrologically corrected to remove all spurious depressions and flat areas. DEM pre-processing is usually achieved using the breach_depressions_least_cost or fill_depressions tool.

In addition to the input DEM/pointer raster name, the user must specify the output type (out_type). The output flow-accumulation can be 1) specific catchment area (SCA), which is the upslope contributing area divided by the contour length (taken as the grid resolution), 2) total catchment area in square-metres, or 3) the number of upslope grid cells. The user must also specify whether the output flow-accumulation grid should be log-tranformed, i.e. the output, if this option is selected, will be the natural-logarithm of the accumulated area. This is a transformation that is often performed to better visualize the contributing area distribution. Because contributing areas tend to be very high along valley bottoms and relatively low on hillslopes, when a flow-accumulation image is displayed, the distribution of values on hillslopes tends to be 'washed out' because the palette is stretched out to represent the highest values. Log-transformation (log) provides a means of compensating for this phenomenon. Importantly, however, log-transformed flow-accumulation grids must not be used to estimate other secondary terrain indices, such as the wetness index, or relative stream power index.

Grid cells possessing the NoData value in the input DEM/pointer raster are assigned the NoData value in the output flow-accumulation image. The output raster is of the float data type and continuous data scale.

Reference

Tarboton, D. G. (1997). A new method for the determination of flow directions and upslope areas in grid digital elevation models. Water resources research, 33(2), 309-319.

See Also

DInfPointer, D8FlowAccumulation, <a href="tool_help.md#quinn_flow_accumulation">quinn_flow_accumulation</a>, <a href="tool_help.md#qin_flow_accumulation">qin_flow_accumulation</a>, FD8FlowAccumulation, MDInfFlowAccumulation`, rho8_pointer, breach_depressions_least_cost, fill_depressions

Function Signature

def dinf_flow_accum(self, dem: Raster, out_type: str = "sca", convergence_threshold: float = float('inf'), log_transform: bool = False, clip: bool = False, input_is_pointer: bool = False) -> Raster: ...

dinf_mass_flux

This tool can be used to perform a mass flux calculation using DEM-based surface flow-routing techniques. For example, it could be used to model the distribution of sediment or phosphorous within a catchment. Flow-routing is based on a D-Infinity flow pointer derived from an input DEM (dem). The user must also specify the names of loading (loading), efficiency (efficiency), and absorption (absorption) rasters, as well as the output raster. Mass Flux operates very much like a flow-accumulation operation except that rather than accumulating catchment areas the algorithm routes a quantity of mass, the spatial distribution of which is specified within the loading image. The efficiency and absorption rasters represent spatial distributions of losses to the accumulation process, the difference being that the efficiency raster is a proportional loss (e.g. only 50% of material within a particular grid cell will be directed downslope) and the absorption raster is an loss specified as a quantity in the same units as the loading image. The efficiency image can range from 0 to 1, or alternatively, can be expressed as a percentage. The equation for determining the mass sent from one grid cell to a neighbouring grid cell is:

Outflowing Mass = (Loading - Absorption + Inflowing Mass) × Efficiency

This tool assumes that each of the three input rasters have the same number of rows and columns and that any NoData cells present are the same among each of the inputs.

See Also

d8_mass_flux

Function Signature

def dinf_mass_flux(self, dem: Raster, loading: Raster, efficiency: Raster, absorption: Raster) -> Raster: ...

dinf_pointer

This tool is used to generate a flow pointer grid (i.e. flow direction) using the D-infinity (Tarboton, 1997) algorithm. Dinf is a multiple-flow-direction (MFD) method because the flow entering each grid cell is routed one or two downslope neighbours, i.e. flow divergence is permitted. The user must specify the name of a digital elevation model (DEM; dem) that has been hydrologically corrected to remove all spurious depressions and flat areas (breach_depressions_least_cost, fill_depressions). DEM pre-processing is usually achieved using the breach_depressions_least_cost or fill_depressions tool1. Flow directions are specified in the output flow-pointer grid (output) as azimuth degrees measured from north, i.e. any value between 0 and 360 degrees is possible. A pointer value of -1 is used to designate a grid cell with no flow-pointer. This occurs when a grid cell has no downslope neighbour, i.e. a pit cell or topographic depression. Like aspect grids, Dinf flow-pointer grids are best visualized using a circular greyscale palette.

Grid cells possessing the NoData value in the input DEM are assigned the NoData value in the output image. The output raster is of the float data type and continuous data scale.

Reference

Tarboton, D. G. (1997). A new method for the determination of flow directions and upslope areas in grid digital elevation models. Water resources research, 33(2), 309-319.

See Also

DInfFlowAccumulation, breach_depressions_least_cost, fill_depressions

Function Signature

def dinf_pointer(self, dem: Raster) -> Raster: ...

direct_decorrelation_stretch

The Direct Decorrelation Stretch (DDS) is a simple type of saturation stretch. The stretch is applied to a colour composite image and is used to improve the saturation, or colourfulness, of the image. The DDS operates by reducing the achromatic (grey) component of a pixel's colour by a scale factor (k), such that the red (r), green (g), and blue (b) components of the output colour are defined as:

rk = r - k min(r, g, b)

gk = g - k min(r, g, b)

bk = b - k min(r, g, b)

The achromatic factor (k) can range between 0 (no effect) and 1 (full saturation stretch), although typical values range from 0.3 to 0.7. A linear stretch is used afterwards to adjust overall image brightness. Liu and Moore (1996) recommend applying a colour balance stretch, such as balance_contrast_enhancement before using the DDS.

Reference

Liu, J.G., and Moore, J. (1996) Direct decorrelation stretch technique for RGB colour composition. International Journal of Remote Sensing, 17:5, 1005-1018.

See Also

create_colour_composite, balance_contrast_enhancement

Function Signature

def direct_decorrelation_stretch(self, image: Raster, achromatic_factor: float = 0.5, clip_percent: float = 1.0) -> Raster: ...

directional_relief

This tool calculates the relief for each grid cell in a digital elevation model (DEM) in a specified direction. Directional relief is an index of the degree to which a DEM grid cell is higher or lower than its surroundings. It is calculated by subtracting the elevation of a DEM grid cell from the average elevation of those cells which lie between it and the edge of the DEM in a specified compass direction. Thus, positive values indicate that a grid cell is lower than the average elevation of the grid cells in a specific direction (i.e. relatively sheltered), whereas a negative directional relief indicates that the grid cell is higher (i.e. relatively exposed). The algorithm is based on a modification of the procedure described by Lapen and Martz (1993). The modifications include: (1) the ability to specify any direction between 0-degrees and 360-degrees (azimuth), and (2) the ability to use a distance-limited search (max_dist), such that the ray-tracing procedure terminates before the DEM edge is reached for longer search paths. The algorithm works by tracing a ray from each grid cell in the direction of interest and evaluating the average elevation along the ray. Linear interpolation is used to estimate the elevation of the surface where a ray does not intersect the DEM grid precisely at one of its nodes. The user must input a DEM raster file (dem) and a hypothetical wind direction. Furthermore, the user is able to constrain the maximum search distance for the ray tracing. If no maximum search distance is specified, each ray will be traced to the edge of the DEM. The units of the output image are the same as the input DEM.

Ray-tracing is a highly computationally intensive task and therefore this tool may take considerable time to operate for larger sized DEMs. This tool is parallelized to aid with computational efficiency. NoData valued grid cells in the input image will be assigned NoData values in the output image. The output raster is of the float data type and continuous data scale. Directional relief is best displayed using the blue-white-red bipolar palette to distinguish between the positive and negative values that are present in the output.

Reference

Lapen, D. R., & Martz, L. W. (1993). The measurement of two simple topographic indices of wind sheltering-exposure from raster digital elevation models. Computers & Geosciences, 19(6), 769-779.

See Also

fetch_analysis, horizon_angle, relative_aspect

Function Signature

def directional_relief(self, dem: Raster, azimuth: float = 0.0, max_dist: float = float('inf')) -> Raster: ...

dissolve

This tool can be used to remove the interior, or shared, boundaries within a vector polygon coverage. You can either dissolve all interior boundaries or dissolve those boundaries along polygons with the same value of a user-specified attribute within the vector's attribute table. It may be desirable to use the VectorCleaning tool to correct any topological errors resulting from the slight misalignment of nodes along shared boundaries in the vector coverage before performing the dissolve operation.

See Also

clip, erase, polygonize

Function Signature

def dissolve(self, input: Vector, dissolve_field: str = "", snap_tolerance: float = 2.220446049250313e-16) -> Vector: ...

distance_to_outlet

Description

This tool calculates the distance of stream grid cells to the channel network outlet cell for each grid cell belonging to a raster stream network. The user must input a raster containing streams data (streams_raster), where stream grid cells are denoted by all positive non-zero values, and a D8 flow pointer (i.e. flow direction) raster (d8_pointer). The pointer image is used to traverse the stream network and must only be created using the D8 algorithm. Stream cells are designated in the streams image as all grid cells with values greater than zero. Thus, all non-stream or background grid cells are commonly assigned either zeros or NoData values. Background cells will be assigned the NoData value in the output image, unless the zero_background parameter is True, in which case non-stream cells will be assigned zero values in the output.

By default, the pointer raster is assumed to use the clockwise indexing method used by Whitebox. If the pointer file contains ESRI flow direction values instead, the esri_pointer parameter must be True.

See Also

downslope_distance_to_stream, length_of_upstream_channels

Parameters

d8_pointer (Raster): The D8 pointer (flow direction) raster.

streams_raster (Raster): The raster object containing the streams data.

esri_pointer (bool): Determines whether the d8_pointer raster contains pointer data in the Esri format. Default is False.

zero_background (bool): Determines whether the background value in the output raster are assigned zero (True) or NoData values (False). Default is False.

Returns

Raster: returning value

Function Signature

def distance_to_outlet(self, d8_pointer: Raster, streams_raster: Raster, esri_pointer: bool = False, zero_background: bool = False) -> Raster: ...

diversity_filter

This tool assigns each cell in the output grid the number of different values in a moving window centred on each grid cell in the input raster. The input image should contain integer values but floating point data are allowable and will be handled by multiplying pixel values by 1000 and rounding. Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values, e.g. 3, 5, 7, 9... If the kernel filter size is the same in the x and y dimensions, the silent filter flag may be used instead (command-line interface only).

See Also

majority_filter

Function Signature

def diversity_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

downslope_distance_to_stream

This tool can be used to calculate the distance from each grid cell in a raster to the nearest stream cell, measured along the downslope flowpath. The user must specify the name of an input digital elevation model (dem) and streams raster (streams). The DEM must have been pre-processed to remove artifact topographic depressions and flat areas (see breach_depressions_least_cost). The streams raster should have been created using one of the DEM-based stream mapping methods, i.e. contributing area thresholding. Stream cells are designated in this raster as all non-zero values. The output of this tool, along with the elevation_above_stream tool, can be useful for preliminary flood plain mapping when combined with high-accuracy DEM data.

By default, this tool calculates flow-path using the D8 flow algorithm. However, the user may specify (dinf) that the tool should use the D-infinity algorithm instead.

See Also

elevation_above_stream, distance_to_outlet

Function Signature

def downslope_distance_to_stream(self, dem: Raster, streams: Raster, use_dinf: bool = False) -> Raster: ...

downslope_flowpath_length

This tool can be used to calculate the downslope flowpath length from each grid cell in a raster to an outlet cell either at the edge of the grid or at the outlet point of a watershed. The user must specify the name of a flow pointer grid (d8_pntr) derived using the D8 flow algorithm (d8_pointer). This grid should be derived from a digital elevation model (DEM) that has been pre-processed to remove artifact topographic depressions and flat areas (breach_depressions_least_cost, fill_depressions). The user may also optionally provide watershed (watersheds) and weights (weights) images. The optional watershed image can be used to define one or more irregular-shaped watershed boundaries. Flowpath lengths are measured within each watershed in the watershed image (each defined by a unique identifying number) as the flowpath length to the watershed's outlet cell.

The optional weight image is multiplied by the flow-length through each grid cell. This can be useful when there is a need to convert the units of the output image. For example, the default unit of flowpath lengths is the same as the input image(s). Thus, if the input image has X-Y coordinates measured in metres, the output image will likely contain very large values. A weight image containing a value of 0.001 for each grid cell will effectively convert the output flowpath lengths into kilometres. The weight image can also be used to convert the flowpath distances into travel times by multiplying the flow distance through a grid cell by the average velocity.

NoData valued grid cells in any of the input images will be assigned NoData values in the output image. The output raster is of the float data type and continuous data scale.

See Also

d8_pointer, elevation_above_stream, breach_depressions_least_cost, fill_depressions, watershed

Function Signature

def downslope_flowpath_length(self, d8_pointer: Raster, watersheds: Raster, weights: Raster, esri_pntr: bool = False) -> Raster: ...

downslope_index

This tool can be used to calculate the downslope index described by Hjerdt et al. (2004). The downslope index is a measure of the slope gradient between a grid cell and some downslope location (along the flowpath passing through the upslope grid cell) that represents a specified vertical drop (i.e. a potential head drop). The index has been shown to be useful for hydrological, geomorphological, and biogeochemical applications.

The user must input a digital elevaton model (DEM) raster. This DEM should be have been pre-processed to remove artifact topographic depressions and flat areas. The user must also specify the head potential drop (d), and the output type. The output type can be either 'tangent', 'degrees', 'radians', or 'distance'. If 'distance' is selected as the output type, the output grid actually represents the downslope flowpath length required to drop d meters from each grid cell. Linear interpolation is used when the specified drop value is encountered between two adjacent grid cells along a flowpath traverse.

Notice that this algorithm is affected by edge contamination. That is, for some grid cells, the edge of the grid will be encountered along a flowpath traverse before the specified vertical drop occurs. In these cases, the value of the downslope index is approximated by replacing d with the actual elevation drop observed along the flowpath. To avoid this problem, the entire watershed containing an area of interest should be contained in the DEM.

Grid cells containing NoData values in any of the input images are assigned the NoData value in the output raster. The output raster is of the float data type and continuous data scale.

Reference

Hjerdt, K.N., McDonnell, J.J., Seibert, J. Rodhe, A. (2004) A new topographic index to quantify downslope controls on local drainage, Water Resources Research, 40, W05602, doi:10.1029/2004WR003130.

Function Signature

def downslope_index(self, dem: Raster, vertical_drop: float, output_type: str = "tangent") -> Raster: ...

edge_contamination

This tool identifs grid cells in a DEM for which the upslope area extends beyond the raster data extent, so-called 'edge-contamined cells'. If a significant number of edge contaminated cells intersect with your area of interest, it is likely that any estimate of upslope area (i.e. flow accumulation) will be under-estimated.

The user must specify the name (dem) of the input digital elevation model (DEM) and the output file (output). The DEM must have been hydrologically corrected to remove all spurious depressions and flat areas. DEM pre-processing is usually achieved using either the breach_depressions_least_cost (also breach_depressions_least_cost) or fill_depressions tool.

Additionally, the user must specify the type of flow algorithm used for the analysis (-flow_type), which must be one of 'd8', 'mfd', or 'dinf', based on each of the D8FlowAccumulation, FD8FlowAccumulation, DInfFlowAccumulation methods respectively.

See Also

D8FlowAccumulation, FD8FlowAccumulation, DInfFlowAccumulation

Function Signature

def edge_contamination(self, dem: Raster, flow_type: str = "mfd", z_factor: float = -1.0) -> Raster: ...

edge_density

This tool calculates the density of edges, or breaks-in-slope within an input digital elevation model (DEM). A break-in-slope occurs between two neighbouring grid cells if the angular difference between their normal vectors is greater than a user-specified threshold value (norm_diff). edge_density calculates the proportion of edge cells within the neighbouring window, of square filter dimension filter, surrounding each grid cell. Therefore, EdgeDensity is a measure of how complex the topographic surface is within a local neighbourhood. It is therefore a measure of topographic texture. It will take a value near 0.0 for smooth sites and 1.0 in areas of high surface roughness or complex topography.

The distribution of edge_density is highly dependent upon the value of the norm_diff used in the calculation. This threshold may require experimentation to find an appropriate value and is likely dependent upon the topography and source data. Nonetheless, experience has shown that edge_density provides one of the best measures of surface texture of any of the available roughness tools.

See Also

circular_variance_of_aspect, multiscale_roughness, surface_area_ratio, ruggedness_index

Function Signature

def edge_density(self, dem: Raster, filter_size: int = 11, normal_diff_threshold: float = 5.0, z_factor: float = 1.0) -> Raster: ...

edge_preserving_mean_filter

This tool performs a type of edge-preserving mean filter operation on an input image (input). The filter, a type of low-pass filter, can be used to emphasize the longer-range variability in an image, effectively acting to smooth the image and to reduce noise in the image. The algorithm calculates the average value in a moving window centred on each grid cell, including in the averaging only the set of neighbouring values for which the absolute value difference with the centre value is less than a specified threshold value (threshold). It is, therefore, similar to the bilateral_filter, except all neighbours within the threshold difference are equally weighted and neighbour distance is not accounted for. Filter kernels are always square, and filter size, is specified using the filter parameter. This dimensions should be odd, positive integer values, e.g. 3, 5, 7, 9...

This tool works with both greyscale and red-green-blue (RGB) input images. RGB images are decomposed into intensity-hue-saturation (IHS) and the filter is applied to the intensity channel. If an RGB image is input, the threshold value must be in the range 0.0-1.0 (more likely less than 0.15), where a value of 1.0 would result in an ordinary mean filter (mean_filter). NoData values in the input image are ignored during filtering.

See Also

mean_filter, bilateral_filter, edge_preserving_mean_filter, gaussian_filter, median_filter, rgb_to_ihs

Function Signature

def edge_preserving_mean_filter(self, raster: Raster, filter_size: int = 11, threshold: float = 15.0) -> Raster: ...

edge_proportion

This tool will measure the edge proportion, i.e. the proportion of grid cells in a patch that are located along the patch's boundary, for an input raster image (input). Edge proportion is an indicator of polygon shape complexity and elongation. The user must specify the name of the output raster file (output), which will be raster layer containing the input features assigned the edge proportion. The user may also optionally choose to output text data for easy input to a spreadsheet or database.

Objects in the input raster are designated by their unique identifiers. Identifier values must be positive, non-zero whole numbers.

See Also

shape_complexity_index_raster, linearity_index, elongation_ratio

Function Signature

def edge_proportion(self, raster: Raster) -> Tuple[Raster, str]: ...

elev_relative_to_min_max

This tool can be used to express the elevation of a grid cell in a digital elevation model (DEM) as a percentage of the relief between the DEM minimum and maximum values. As such, it provides a basic measure of relative topographic position.

See Also

elev_relative_to_watershed_min_max, elevation_above_stream, ElevAbovePit

Function Signature

def elev_relative_to_min_max(self, dem: Raster) -> Raster: ...

elev_relative_to_watershed_min_max

This tool can be used to express the elevation of a grid cell in a digital elevation model (DEM) as a percentage of the relief between the watershed minimum and maximum values. As such, it provides a basic measure of relative topographic position. The user must input a DEM (dem) and watersheds (watersheds) raster files.

See Also

elev_relative_to_min_max, elevation_above_stream, ElevAbovePit

Function Signature

def elev_relative_to_watershed_min_max(self, dem: Raster, watersheds: Raster) -> Raster: ...

elevation_above_pit

This tool will calculate the elevation of each grid cell in a digital elevation model (DEM) above the nearest downslope pit cell or grid edge cell, depending on which is encountered first during the flow-path traverse. The resulting image is therefore a measure of relative landscape position. The user must input a D8 flow pointer grid and a DEM file. The flow pointer grid must be derived using the D8 flow algorithm.

See Also

elevation_above_stream

Function Signature

def elevation_above_pit(self, dem: Raster) -> Raster: ...

elevation_above_stream

This tool can be used to calculate the elevation of each grid cell in a raster above the nearest stream cell, measured along the downslope flowpath. This terrain index, a measure of relative topographic position, is essentially equivalent to the 'height above drainage' (HAND), as described by Renno et al. (2008). The user must specify the name of an input digital elevation model (dem) and streams raster (streams). The DEM must have been pre-processed to remove artifact topographic depressions and flat areas (see breach_depressions_least_cost). The streams raster should have been created using one of the DEM-based stream mapping methods, i.e. contributing area thresholding. Stream cells are designated in this raster as all non-zero values. The output of this tool, along with the downslope_distance_to_stream tool, can be useful for preliminary flood plain mapping when combined with high-accuracy DEM data.

The difference between elevation_above_stream and elevation_above_stream_euclidean is that the former calculates distances along drainage flow-paths while the latter calculates straight-line distances to streams channels.

Reference

Renno, C. D., Nobre, A. D., Cuartas, L. A., Soares, J. V., Hodnett, M. G., Tomasella, J., & Waterloo, M. J. (2008). HAND, a new terrain descriptor using SRTM-DEM: Mapping terra-firme rainforest environments in Amazonia. Remote Sensing of Environment, 112(9), 3469-3481.

See Also

elevation_above_stream_euclidean, downslope_distance_to_stream, ElevAbovePit, breach_depressions_least_cost

Function Signature

def elevation_above_stream(self, dem: Raster, streams: Raster) -> Raster: ...

elevation_above_stream_euclidean

This tool can be used to calculate the elevation of each grid cell in a raster above the nearest stream cell, measured along the straight-line distance. This terrain index, a measure of relative topographic position, is related to the 'height above drainage' (HAND), as described by Renno et al. (2008). HAND is generally estimated with distances measured along drainage flow-paths, which can be calculated using the elevation_above_stream tool. The user must specify the name of an input digital elevation model (dem) and streams raster (streams). Stream cells are designated in this raster as all non-zero values. The output of this tool, along with the downslope_distance_to_stream tool, can be useful for preliminary flood plain mapping when combined with high-accuracy DEM data.

The difference between elevation_above_stream and elevation_above_stream_euclidean is that the former calculates distances along drainage flow-paths while the latter calculates straight-line distances to streams channels.

Reference

Renno, C. D., Nobre, A. D., Cuartas, L. A., Soares, J. V., Hodnett, M. G., Tomasella, J., & Waterloo, M. J. (2008). HAND, a new terrain descriptor using SRTM-DEM: Mapping terra-firme rainforest environments in Amazonia. Remote Sensing of Environment, 112(9), 3469-3481.

See Also

elevation_above_stream, downslope_distance_to_stream, ElevAbovePit

Function Signature

def elevation_above_stream_euclidean(self, dem: Raster, streams: Raster) -> Raster: ...

elevation_percentile

Elevation percentile (EP) is a measure of local topographic position (LTP). It expresses the vertical position for a digital elevation model (DEM) grid cell (z0) as the percentile of the elevation distribution within the filter window, such that:

EP = counti∈C(zi > z0) x (100 / nC)

where z0 is the elevation of the window's center grid cell, zi is the elevation of cell i contained within the neighboring set C, and nC is the number of grid cells contained within the window.

EP is unsigned and expressed as a percentage, bound between 0% and 100%. Quantile-based estimates (e.g., the median and interquartile range) are often used in nonparametric statistics to provide data variability estimates without assuming the distribution is normal. Thus, EP is largely unaffected by irregularly shaped elevation frequency distributions or by outliers in the DEM, resulting in a highly robust metric of LTP. In fact, elevation distributions within small to medium sized neighborhoods often exhibit skewed, multimodal, and non-Gaussian distributions, where the occurrence of elevation errors can often result in distribution outliers. Thus, based on these statistical characteristics, EP is considered one of the most robust representation of LTP.

The algorithm implemented by this tool uses the relatively efficient running-histogram filtering algorithm of Huang et al. (1979). Because most DEMs contain floating point data, elevation values must be rounded to be binned. The sig_digits parameter is used to determine the level of precision preserved during this binning process. The algorithm is parallelized to further aid with computational efficiency.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

References

Newman, D. R., Lindsay, J. B., and Cockburn, J. M. H. (2018). Evaluating metrics of local topographic position for multiscale geomorphometric analysis. Geomorphology, 312, 40-50.

Huang, T., Yang, G.J.T.G.Y. and Tang, G., 1979. A fast two-dimensional median filtering algorithm. IEEE Transactions on Acoustics, Speech, and Signal Processing, 27(1), pp.13-18.

See Also

DevFromMeanElev, DiffFromMeanElev

Function Signature

def elevation_percentile(self, dem: Raster, filter_size_x: int = 11, filter_size_y: int = 11, sig_digits: int = 2) -> Raster: ...

eliminate_coincident_points

This tool can be used to remove any coincident, or nearly coincident, points from a vector points file. The user must specify the name of the input file, which must be of a POINTS VectorGeometryType, the output file name, and the tolerance distance. All points that are within the specified tolerance distance will be eliminated from the output file. A tolerance distance of 0.0 indicates that points must be exactly coincident to be removed.

See Also

LidarRemoveDuplicates

Function Signature

def eliminate_coincident_points(self, input: Vector, tolerance_dist: float) -> Vector: ...

elongation_ratio

This tool can be used to calculate the elongation ratio for vector polygons. The elongation ratio values calculated for each vector polygon feature will be placed in the accompanying database file (.dbf) as an elongation field (ELONGATION).

The elongation ratio (E) is:

E = 1 - S / L

Where S is the short-axis length, and L is the long-axis length. Axes lengths are determined by estimating the minimum bounding box.

The elongation ratio provides similar information as the Linearity Index. The ratio is not an adequate measure of overall polygon narrowness, because a highly sinuous but narrow polygon will have a low linearity (elongation) owing to the compact nature of these polygon.

Function Signature

def elongation_ratio(self, input: Vector) -> Vector: ...

embankment_mapping

This tool can be used to map and/or remove road embankments from an input fine-resolution digital elevation model (dem). Fine-resolution LiDAR DEMs can represent surface features such as road and railway embankments with high fidelity. However, transportation embankments are problematic for several environmental modelling applications, including soil an vegetation distribution mapping, where the pre-embankment topography is the contolling factor. The algorithm utilizes repositioned (search_dist) transportation network cells, derived from rasterizing a transportation vector (road_vec), as seed points in a region-growing operation. The embankment region grows based on derived morphometric parameters, including road surface width (min_road_width), embankment width (typical_width and max_width), embankment height (max_height), and absolute slope (spillout_slope). The tool can be run in two modes. By default the tool will simply map embankment cells, with a Boolean output raster. If, however, the remove_embankments flag is specified, the tool will instead output a DEM for which the mapped embankment grid cells have been excluded and new surfaces have been interpolated based on the surrounding elevation values (see below).

Hillshade from original DEM:

Hillshade from embankment-removed DEM:

References

Van Nieuwenhuizen, N, Lindsay, JB, DeVries, B. 2021. Automated mapping of transportation embankments in fine-resolution LiDAR DEMs. Remote Sensing. 13(7), 1308; https://doi.org/10.3390/rs13071308

See Also:

remove_off_terrain_objects, SmoothVegetationResidual

Function Signature

def embankment_mapping(self, dem: Raster, roads_vector: Vector, search_dist: float = 2.5, min_road_width: float = 6.0, typical_embankment_width: float = 30.0, typical_embankment_max_height: float = 2.0, embankment_max_width: float = 60.0, max_upwards_increment: float = 0.05, spillout_slope: float = 4.0, remove_embankments: bool = False) -> Tuple[Raster, Union[Raster, None]]: ...

emboss_filter

This tool can be used to perform one of eight 3x3 emboss filters on a raster image. Like the sobel_filter and prewitt_filter, the emboss_filter is often applied in edge-detection applications. While these other two common edge-detection filters approximate the slope magnitude of the local neighbourhood surrounding each grid cell, the emboss_filter can be used to estimate the directional slope. The kernel weights for each of the eight available filters are as follows:

North (n)

...
0-10
000
010

Northeast (ne)

...
00-1
000
-100

East (e)

...
000
10-1
000

Southeast (se)

...
100
000
00-1

South (s)

...
010
100
0-10

Southwest (sw)

...
001
000
-100

West (w)

...
000
-101
000

Northwest (nw)

...
-100
000
001

The user must specify the direction, options include 'n', 's', 'e', 'w', 'ne', 'se', 'nw', 'sw'. The user may also optionally clip the output image distribution tails by a specified amount (e.g. 1%).

See Also

sobel_filter, prewitt_filter

Function Signature

def emboss_filter(self, raster: Raster, direction: str = "n", clip_amount: float = 0.0) -> Raster: ...

erase

This tool will remove all the features, or parts of features, that overlap with the features of the erase vector file. The erasing operation is one of the most common vector overlay operations in GIS and effectively imposes the boundary of the erase layer on a set of input vector features, or target features.

See Also

clip

Function Signature

def erase(self, input: Vector, erase_layer: Vector) -> Vector: ...

erase_polygon_from_lidar

This tool can be used to isolate, or clip, all of the LiDAR points in a LAS file (input) contained within one or more vector polygon features. The user must specify the name of the input clip file (--polygons), which must be a vector of a Polygon base shape type. The clip file may contain multiple polygon features and polygon hole parts will be respected during clipping, i.e. LiDAR points within polygon holes will be removed from the output LAS file.

Use the erase_polygon_from_lidar tool to perform the complementary operation of removing points from a LAS file that are contained within a set of polygons.

See Also

erase_polygon_from_lidar, FilterLidar, clip, clip_raster_to_polygon

Function Signature

def erase_polygon_from_lidar(self, input: Lidar, polygons: Vector) -> Lidar: ...

erase_polygon_from_raster

This tool can be used to set values an input raster (input) to a NoData background value with a vector erasing polygon (polygons). The input erase polygon file must be a vector of a Polygon base shape type. The erase file may contain multiple polygon features. Polygon hole parts will be respected during clipping, i.e. polygon holes will not be removed from the output raster. Raster grid cells that fall inside of a polygons in the erase file will be assigned the NoData background value in the output file.

See Also

clip_raster_to_polygon

Function Signature

def erase_polygon_from_raster(self, raster: Raster, polygons: Vector) -> Raster: ...

euclidean_allocation

This tool assigns grid cells in the output image the value of the nearest target cell in the input image, measured by the Euclidean distance (i.e. straight-line distance). Thus, euclidean_allocation essentially creates the Voronoi diagram for a set of target cells. Target cells are all non-zero, non-NoData grid cells in the input image. Distances are calculated using the same efficient algorithm (Shih and Wu, 2003) as the euclidean_distance tool.

Reference

Shih FY and Wu Y-T (2004), Fast Euclidean distance transformation in two scans using a 3 x 3 neighborhood, Computer Vision and Image Understanding, 93: 195-205.

See Also

euclidean_distance, voronoi_diagram, cost_allocation

Function Signature

def euclidean_allocation(self, input: Raster) -> Raster: ...

euclidean_distance

This tool will estimate the Euclidean distance (i.e. straight-line distance) between each grid cell and the nearest 'target cell' in the input image. Target cells are all non-zero, non-NoData grid cells. Distance in the output image is measured in the same units as the horizontal units of the input image.

Algorithm Description

The algorithm is based on the highly efficient distance transform of Shih and Wu (2003). It makes four passes of the image; the first pass initializes the output image; the second and third passes calculate the minimum squared Euclidean distance by examining the 3 x 3 neighbourhood surrounding each cell; the last pass takes the square root of cell values, transforming them into true Euclidean distances, and deals with NoData values that may be present. All NoData value grid cells in the input image will contain NoData values in the output image. As such, NoData is not a suitable background value for non-target cells. Background areas should be designated with zero values.

Reference

Shih FY and Wu Y-T (2004), Fast Euclidean distance transformation in two scans using a 3 x 3 neighborhood, Computer Vision and Image Understanding, 93: 195-205.

See Also

euclidean_allocation, cost_distance

Function Signature

def euclidean_distance(self, input: Raster) -> Raster: ...

export_table_to_csv

This tool can be used to export a vector's attribute table to a comma separated values (CSV) file. CSV files stores tabular data (numbers and text) in plain-text form such that each row corresponds to a record and each column to a field. Fields are typically separated by commas within records. The user must specify the name of the vector (and associated attribute file), the name of the output CSV file, and whether or not to include the field names as a header column in the output CSV file.

See Also

merge_table_with_csv

Function Signature

def export_table_to_csv(self, input: Vector, output_csv_file: str, headers: bool = True) -> None: ...

exposure_towards_wind_flux

This tool creates a new raster in which each grid cell is assigned the exposure of the land-surface to a hypothetical wind flux. It can be conceptualized as the angle between a plane orthogonal to the wind and a plane that represents the local topography at a grid cell (Bohner and Antonic, 2007). The user must input a digital elevation model (dem), as well as the dominant wind azimuth (azimuth) and a maximum search distance (max_dist) used to calclate the horizon angle. Notice that the specified azimuth represents a regional average wind direction.

Exposure towards the sloped wind flux essentially combines the relative terrain aspect and the maximum upwind slope (i.e. horizon angle). This terrain attribute accounts for land-surface orientation, relative to the wind, and shadowing effects of distant topographic features but does not account for deflection of the wind by topography. This tool should not be used on very extensive areas over which Earth's curvature must be taken into account. DEMs in projected coordinate systems are preferred.

Algorithm Description:

Exposure is measured based on the equation presented in Antonic and Legovic (1999):

cos(E) = cos(S) sin(H) + sin(S) cos(H) cos(Az - A)

Where, E is angle between a plane defining the local terrain and a plane orthogonal to the wind flux, S is the terrain slope, A is the terrain aspect, Az is the azimuth of the wind flux, and H is the horizon angle of the wind flux, which is zero when only the horizontal component of the wind flux is accounted for.

Exposure images are best displayed using a greyscale or bipolar palette to distinguish between the positive and negative values that are present in the output.

References

Antonić, O., & Legović, T. 1999. Estimating the direction of an unknown air pollution source using a digital elevation model and a sample of deposition. Ecological modelling, 124(1), 85-95.

Böhner, J., & Antonić, O. 2009. Land-surface parameters specific to topo-climatology. Developments in Soil Science, 33, 195-226.

See Also

relative_aspect

Function Signature

def exposure_towards_wind_flux(self, dem: Raster, azimuth: float = 0.0, max_dist: float = float('inf'), z_factor: float = 1.0) -> Raster: ...

extend_vector_lines

This tool can be used to extend vector lines by a specified distance. The user must input the names of the input and output shapefiles, the distance to extend features by, and whether to extend both ends, line starts, or line ends. The input shapefile must be of a POLYLINE base shape type and should be in a projected coordinate system.

Function Signature

def extend_vector_lines(self, input: Vector, distance: float, extend_direction: str = "both") -> Vector: ...

extract_nodes

This tool converts vector lines or polygons into vertex points. The user must specify the name of the input vector, which must be of a polyline or polygon base shape type, and the name of the output point-type vector.

Function Signature

def extract_nodes(self, input: Vector) -> Vector: ...

extract_raster_values_at_points

This tool can be used to extract the values of one or more rasters (inputs) at the sites of a set of vector points. By default, the data is output to the attribute table of the input points (points) vector; however, if the out_text parameter is specified, the tool will additionally output point values as text data to standard output (stdout). Attribute fields will be added to the table of the points file, with field names, VALUE1, VALUE2, VALUE3, etc. each corresponding to the order of input rasters.

If you need to plot a chart of values from a raster stack at a set of points, the image_stack_profile may be more suitable for this application.

See Also

image_stack_profile, find_lowest_or_highest_points

Function Signature

def extract_raster_values_at_points(self, rasters: List[Raster], points: Vector) -> Tuple[Vector, str]: ...

extract_streams

Description

This tool can be used to extract, or map, the likely stream cells from an input flow-accumulation image (flow_accumulation). The algorithm applies a threshold to the input flow accumulation image such that streams are considered to be all grid cells with accumulation values greater than the specified threshold (threshold). As such, this threshold represents the minimum area (area is used here as a surrogate for discharge) required to initiate and maintain a channel. Smaller threshold values result in more extensive stream networks and vice versa. Unfortunately there is very little guidance regarding an appropriate method for determining the channel initiation area threshold in practice. As such, it is frequently determined either by examining map or imagery data, using field work, or by experimentation until a suitable or desirable channel network is identified. Notice that the threshold value will be unique for each landscape and dataset (including source and grid resolution), further complicating its a priori determination. There is also evidence that in some landscape the threshold is a combined upslope area-slope function. Generally, a lower threshold is appropriate in humid climates and a higher threshold is appropriate in areas underlain by more resistant bedrock. Climate and bedrock resistance are two factors related to drainage density, i.e. the extent to which a landscape is dissected by drainage channels.

The background value of the output raster will be the NoData value unless zero_background is set to True.

See Also

extract_valleys

Parameters

flow_accumulation (Raster): The input flow accumulation Raster object.

threshold (float): The minimum accumulation value required to be part of a stream channel. Default is 0.0, but should be set higher.

zero_background (bool): Whether the output raster uses 0.0 for non-channel cells (True) or NoData (False). Default is False.

Returns:

Raster

Function Signature

def extract_streams(self, flow_accumulation: Raster, threshold: float = 0.0, zero_background: bool = False) -> Raster: ...

extract_valleys

This tool can be used to extract channel networks from an input digital elevation models (dem) using one of three techniques that are based on local topography alone.

The Lindsay (2006) 'lower-quartile' method (variant='LQ') algorithm is a type of 'valley recognition' method. Other channel mapping methods, such as the Johnston and Rosenfeld (1975) algorithm, experience problems because channel profiles are not always 'v'-shaped, nor are they always apparent in small 3 x 3 windows. The lower-quartile method was developed as an alternative and more flexible valley recognition channel mapping technique. The lower-quartile method operates by running a filter over the DEM that calculates the percentile value of the centre cell with respect to the distribution of elevations within the filter window. The roving window is circular, the diameter of which should reflect the topographic variation of the area (e.g. the channel width or average hillslope length). If this variant is selected, the user must specify the filter_size parameter, in pixels, and this value should be an odd number (e.g. 3, 5, 7, etc.). The appropriateness of the selected window diameter will depend on the grid resolution relative to the scale of topographic features. Cells that are within the lower quartile of the distribution of elevations of their neighbourhood are flagged. Thus, the algorithm identifies grid cells that are in relatively low topographic positions at a local scale. This approach to channel mapping is only appropriate in fluvial landscapes. In regions containing numerous lakes and wetlands, the algorithm will pick out the edges of features.

The Johnston and Rosenfeld (1975) algorithm (variant='JandR') is a type of 'valley recognition' method and operates as follows: channel cells are flagged in a 3 x 3 window if the north and south neighbours are higher than the centre grid cell or if the east and west neighbours meet this same criterion. The group of cells that are flagged after one pass of the roving window constituted the drainage network. This method is best applied to DEMs that are relatively smooth and do not exhibit high levels of short-range roughness. As such, it may be desirable to use a smoothing filter before applying this tool. The feature_preserving_smoothing is a good option for removing DEM roughness while preserving the topographic information contain in breaks-in-slope (i.e. edges).

The Peucker and Douglas (1975) algorithm (variant='PandD') is one of the simplest and earliest algorithms for topography-based network extraction. Their 'valley recognition' method operates by passing a 2 x 2 roving window over a DEM and flagging the highest grid cell in each group of four. Once the window has passed over the entire DEM, channel grid cells are left unflagged. This method is also best applied to DEMs that are relatively smooth and do not exhibit high levels of short-range roughness. Pre-processing the DEM with the feature_preserving_smoothing tool may also be useful when applying this method.

Each of these methods of extracting valley networks result in line networks that can be wider than a single grid cell. As such, it is often desirable to thin the resulting network using a line-thinning algorithm. The option to perform line-thinning is provided by the tool as a post-processing step (line_thin=True).

References

Johnston, E. G., & Rosenfeld, A. (1975). Digital detection of pits, peaks, ridges, and ravines. IEEE Transactions on Systems, Man, and Cybernetics, (4), 472-480.

Lindsay, J. B. (2006). Sensitivity of channel mapping techniques to uncertainty in digital elevation data. International Journal of Geographical Information Science, 20(6), 669-692.

Peucker, T. K., & Douglas, D. H. (1975). Detection of surface-specific points by local parallel processing of discrete terrain elevation data. Computer Graphics and image processing, 4(4), 375-387.

See Also

feature_preserving_smoothing

Function Signature

def extract_valleys(self, dem: Raster, variant: str = "lq", line_thin: bool = False, filter_size: int = 5) -> Raster: ...

farthest_channel_head

Description

This tool calculates the upstream distance to the farthest stream head for each grid cell belonging to a raster stream network. The user must input a raster containing streams data (streams), where stream grid cells are denoted by all positive non-zero values, and a D8 flow pointer (i.e. flow direction) raster (d8_pointer). The pointer image is used to traverse the stream network and must only be created using the D8 algorithm. Stream cells are designated in the streams image as all values greater than zero. Thus, all non-stream or background grid cells are commonly assigned either zeros or NoData values. Background cells will be assigned the NoData value in the output image, unless zero_background=True, in which case non-stream cells will be assigned zero values in the output.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, the user should specify esri_pntr=True.

See Also

length_of_upstream_channels, find_main_stem

Parameters

d8_pointer (Raster): The D8 pointer (flow direction) raster.

streams_raster (Raster): The raster object containing the streams data.

esri_pointer (bool): Determines whether the d8_pointer raster contains pointer data in the Esri format. Default is False.

zero_background (bool): Determines whether the background value in the output raster are assigned zero (True) or NoData values (False). Default is False.

Returns

Raster: returning value

Function Signature

def farthest_channel_head(self, d8_pointer: Raster, streams_raster: Raster, esri_pointer: bool = False, zero_background: bool = False) -> Raster: ...

fast_almost_gaussian_filter

The tool is somewhat modified from Dr. Kovesi's original Matlab code in that it works with both greyscale and RGB images (decomposes to HSI and uses the intensity data) and it handles the case of rasters that contain NoData values. This adds complexity to the original 20 additions and 5 multiplications assertion of the original paper.

Also note, for small values of sigma (< 1.8), you should probably just use the regular GaussianFilter tool.

Reference

P. Kovesi 2010 Fast Almost-Gaussian Filtering, Digital Image Computing: Techniques and Applications (DICTA), 2010 International Conference on.

Function Signature

def fast_almost_gaussian_filter(self, raster: Raster, sigma: float = 1.8) -> Raster: ...

fd8_flow_accum

This tool is used to generate a flow accumulation grid (i.e. contributing area) using the FD8 algorithm (Freeman, 1991), sometimes referred to as FMFD. This algorithm is an examples of a multiple-flow-direction (MFD) method because the flow entering each grid cell is routed to each downslope neighbour, i.e. flow divergence is permitted. The user must specify the name (dem) of the input digital elevation model (DEM). The DEM must have been hydrologically corrected to remove all spurious depressions and flat areas. DEM pre-processing is usually achieved using either the breach_depressions_least_cost (also breach_depressions_least_cost) or fill_depressions tool. A value must also be specified for the exponent parameter (exponent), a number that controls the degree of dispersion in the resulting flow-accumulation grid. A lower value yields greater apparent flow dispersion across divergent hillslopes. Some experimentation suggests that a value of 1.1 is appropriate (Freeman, 1991), although this is almost certainly landscape-dependent.

In addition to the input DEM, the user must specify the output type (out_type). The output flow-accumulation can be 1) cells (i.e. the number of inflowing grid cells), catchment area (i.e. the upslope area), or specific contributing area (i.e. the catchment area divided by the flow width. The default value is cells. The user must also specify whether the output flow-accumulation grid should be log-tranformed (log), i.e. the output, if this option is selected, will be the natural-logarithm of the accumulated flow value. This is a transformation that is often performed to better visualize the contributing area distribution. Because contributing areas tend to be very high along valley bottoms and relatively low on hillslopes, when a flow-accumulation image is displayed, the distribution of values on hillslopes tends to be 'washed out' because the palette is stretched out to represent the highest values. Log-transformation provides a means of compensating for this phenomenon. Importantly, however, log-transformed flow-accumulation grids must not be used to estimate other secondary terrain indices, such as the wetness index, or relative stream power index.

The non-dispersive threshold (threshold) is a flow-accumulation value (measured in upslope grid cells, which is directly proportional to area) above which flow dispersion is no longer permitted. Grid cells with flow-accumulation values above this threshold will have their flow routed in a manner that is similar to the D8 single-flow-direction algorithm, directing all flow towards the steepest downslope neighbour. This is usually done under the assumption that flow dispersion, whilst appropriate on hillslope areas, is not realistic once flow becomes channelized.

Reference

Freeman, T. G. (1991). Calculating catchment area with divergent flow based on a regular grid. Computers and Geosciences, 17(3), 413-422.

See Also

D8FlowAccumulation, quinn_flow_accumulation, qin_flow_accumulation, DInfFlowAccumulation, MDInfFlowAccumulation, rho8_pointer

Function Signature

def fd8_flow_accum(self, dem: Raster, out_type: str = "sca", exponent: float = 1.1, convergence_threshold: float = float('inf'), log_transform: bool = False, clip: bool = False) -> Raster: ...

fd8_pointer

This tool is used to generate a flow pointer grid (i.e. flow direction) using the FD8 (Freeman, 1991) algorithm. FD8 is a multiple-flow-direction (MFD) method because the flow entering each grid cell is routed one or more downslope neighbours, i.e. flow divergence is permitted. The user must specify the name of a digital elevation model (DEM; dem) that has been hydrologically corrected to remove all spurious depressions and flat areas. DEM pre-processing is usually achieved using the breach_depressions_least_cost or fill_depressions tools.

By default, D8 flow pointers use the following clockwise, base-2 numeric index convention:

...
641281
3202
1684

In the case of the FD8 algorithm, some portion of the flow entering a grid cell will be sent to each downslope neighbour. Thus, the FD8 flow-pointer value is the sum of each of the individual pointers for all downslope neighbours. For example, if a grid cell has downslope neighbours to the northeast, east, and south the corresponding FD8 flow-pointer value will be 1 + 2 + 8 = 11. Using the naming convention above, this is the only combination of flow-pointers that will result in the combined value of 11. Using the base-2 naming convention allows for the storage of complex combinations of flow-points using a single numeric value, which is the reason for using this somewhat odd convention.

Reference

Freeman, T. G. (1991). Calculating catchment area with divergent flow based on a regular grid. Computers and Geosciences, 17(3), 413-422.

See Also

FD8FlowAccumulation, d8_pointer, DInfPointer, breach_depressions_least_cost, fill_depressions

Function Signature

def fd8_pointer(self, dem: Raster) -> Raster: ...

feature_preserving_smoothing

Description

This tool implements a highly modified form of the DEM de-noising algorithm described by Sun et al. (2007). It is very effective at removing surface roughness from digital elevation models (DEMs), without significantly altering breaks-in-slope. As such, this tool should be used for smoothing DEMs rather than either smoothing with low-pass filters (e.g. mean, median, Gaussian filters) or grid size coarsening by resampling. The algorithm works by 1) calculating the surface normal 3D vector of each grid cell in the DEM, 2) smoothing the normal vector field using a filtering scheme that applies more weight to neighbours with lower angular difference in surface normal vectors, and 3) uses the smoothed normal vector field to update the elevations in the input DEM.

Sun et al.'s (2007) original method was intended to work on input point clouds and fitted triangular irregular networks (TINs). The algorithm has been modified to work with input raster DEMs instead. In so doing, this algorithm calculates surface normal vectors from the planes fitted to 3 x 3 neighbourhoods surrounding each grid cell, rather than the triangular facet. The normal vector field smoothing and elevation updating procedures are also based on raster filtering operations. These modifications make this tool more efficient than Sun's original method, but will also result in a slightly different output than what would be achieved with Sun's method.

The user must specify the values of three key parameters, including the filter size (filter), the normal difference threshold (norm_diff), and the number of iterations (num_iter). Lindsay et al. (2019) found that the degree of smoothing was less impacted by the filter size than it was either the normal difference threshold and the number of iterations. A filter size of 11, the default value, tends to work well in many cases. To increase the level of smoothing applied to the DEM, consider increasing the normal difference threshold, i.e. the angular difference in normal vectors between the center cell of a filter window and a neighbouring cell. This parameter determines which neighbouring values are included in a filtering operation and higher values will result in a greater number of neighbouring cells included, and therefore smoother surfaces. Similarly, increasing the number of iterations from the default value of 3 to upwards of 5-10 will result in significantly greater smoothing.

Before smoothing treatment:

After smoothing treatment with FPS:

For a video tutorial on how to use the feature_preserving_smoothing tool, please see this YouTube video.

Reference

Lindsay JB, Francioni A, Cockburn JMH. 2019. LiDAR DEM smoothing and the preservation of drainage features. Remote Sensing, 11(16), 1926; DOI: 10.3390/rs11161926.

Sun, X., Rosin, P., Martin, R., & Langbein, F. (2007). Fast and effective feature-preserving mesh denoising. IEEE Transactions on Visualization & Computer Graphics, (5), 925-938.

Parameters

dem (Raster): The input digital elevation model (DEM)

filter_size (int): The filter size used for smoothing. Default is 11.

normal_diff_threshold (float): The maximum allowable difference in the angle of the normals between two grid cells on the same facet. Default is 8.0.

iterations (int): The number of iterations used during smoothing. Default is 3.

max_elevation_diff (float): The maximum allowable vertical distance that a cell's elevation is allowed to be changed by

z_factor (float): Used to convert elevation units so that they match the horizontal units. Unless the two units differ, this should be set to 1.0. Default is 1.0.

Returns

Raster: return value

Function Signature

def feature_preserving_smoothing(self, dem: Raster, filter_size: int = 11, normal_diff_threshold: float = 8.0, iterations: int = 3, max_elevation_diff: float = float('inf'), z_factor: float = 1.0) -> Raster: ...

fetch_analysis

This tool creates a new raster in which each grid cell is assigned the distance, in meters, to the nearest topographic obstacle in a specified direction. It is a modification of the algorithm described by Lapen and Martz (1993). Unlike the original algorithm, Fetch Analysis is capable of analyzing fetch in any direction from 0-360 degrees. The user must input a digital elevation model (DEM) raster file, a hypothetical wind direction, and a value for the height increment parameter. The algorithm searches each grid cell in a path following the specified wind direction until the following condition is met:

Ztest >= Zcore + DI

where Zcore is the elevation of the grid cell at which fetch is being determined, Ztest is the elevation of the grid cell being tested as a topographic obstacle, D is the distance between the two grid cells in meters, and I is the height increment in m/m. Lapen and Martz (1993) suggest values for I in the range of 0.025 m/m to 0.1 m/m based on their study of snow re-distribution in low-relief agricultural landscapes of the Canadian Prairies. If the directional search does not identify an obstacle grid cell before the edge of the DEM is reached, the distance between the DEM edge and Zcore is entered. Edge distances are assigned negative values to differentiate between these artificially truncated fetch values and those for which a valid topographic obstacle was identified. Notice that linear interpolation is used to estimate the elevation of the surface where a ray (i.e. the search path) does not intersect the DEM grid precisely at one of its nodes.

Ray-tracing is a highly computationally intensive task and therefore this tool may take considerable time to operate for larger sized DEMs. This tool is parallelized to aid with computational efficiency. NoData valued grid cells in the input image will be assigned NoData values in the output image. Fetch Analysis images are best displayed using the blue-white-red bipolar palette to distinguish between the positive and negative values that are present in the output.

Reference

Lapen, D. R., & Martz, L. W. (1993). The measurement of two simple topographic indices of wind sheltering-exposure from raster digital elevation models. Computers & Geosciences, 19(6), 769-779.

See Also

directional_relief, horizon_angle, relative_aspect

Function Signature

def fetch_analysis(self, dem: Raster, azimuth: float = 0.0, height_increment: float = 0.05) -> Raster: ...

fill_burn

Burns streams into a DEM using the FillBurn (Saunders, 1999) method which produces a hydro-enforced DEM. This tool uses the algorithm described in:

Lindsay JB. 2016. The practice of DEM stream burning revisited. Earth Surface Processes and Landforms, 41(5): 658-668. DOI: 10.1002/esp.3888

And:

Saunders, W. 1999. Preparation of DEMs for use in environmental modeling analysis, in: ESRI User Conference. pp. 24-30.

Function Signature

def fill_burn(self, dem: Raster, streams: Vector) -> Raster: ...

fill_depressions

This tool can be used to fill all of the depressions in a digital elevation model (DEM) and to remove the flat areas. This is a common pre-processing step required by many flow-path analysis tools to ensure continuous flow from each grid cell to an outlet located along the grid edge. The fill_depressions algorithm operates by first identifying single-cell pits, that is, interior grid cells with no lower neighbouring cells. Each pit cell is then visited from highest to lowest and a priority region-growing operation is initiated. The area of monotonically increasing elevation, starting from the pit cell and growing based on flood order, is identified. Once a cell, that has not been previously visited and possessing a lower elevation than its discovering neighbour cell, is identified the discovering neighbour is labelled as an outlet (spill point) and the outlet elevation is noted. The algorithm then back-fills the labelled region, raising the elevation in the output DEM (output) to that of the outlet. Once this process is completed for each pit cell (noting that nested pit cells are often solved by prior pits) the flat regions of filled pits are optionally treated (fix_flats) with an applied small slope gradient away from outlets (note, more than one outlet cell may exist for each depression). The user may optionally specify the size of the elevation increment used to solve flats (flat_increment), although it is best to not specify this optional value and to let the algorithm determine the most suitable value itself. The flat-fixing method applies a small gradient away from outlets using another priority region-growing operation (i.e. based on a priority queue operation), where priorities are set by the elevations in the input DEM (input). This in effect ensures a gradient away from outlet cells but also following the natural pre-conditioned topography internal to depression areas. For example, if a large filled area occurs upstream of a damming road-embankment, the filled DEM will possess flow directions that are similar to the un-flooded valley, with flow following the valley bottom. In fact, the above case is better handled using the breach_depressions_least_cost tool, which would simply cut through the road embankment at the likely site of a culvert. However, the flat-fixing method of fill_depressions does mean that this common occurrence in LiDAR DEMs is less problematic.

The breach_depressions_least_cost, while slightly less efficient than either other hydrological preprocessing methods, often provides a lower impact solution to topographic depressions and should be preferred in most applications. In comparison with the breach_depressions_least_cost tool, the depression filling method often provides a less satisfactory, higher impact solution. It is advisable that users try the breach_depressions_least_cost tool to remove depressions from their DEMs before using fill_depressions. Nonetheless, there are applications for which full depression filling using the fill_depressions tool may be preferred.

Note that this tool will not fill in NoData regions within the DEM. It is advisable to remove such regions using the fill_missing_data tool prior to application.

See Also

breach_depressions_least_cost, breach_depressions_least_cost, sink, depth_in_sink, fill_missing_data

Function Signature

def fill_depressions(self, dem: Raster, fix_flats: bool = True, flat_increment: float = float('nan'), max_depth: float = float('inf')) -> Raster: ...

fill_depressions_planchon_and_darboux

This tool can be used to fill all of the depressions in a digital elevation model (DEM) and to remove the flat areas using the Planchon and Darboux (2002) method. This is a common pre-processing step required by many flow-path analysis tools to ensure continuous flow from each grid cell to an outlet located along the grid edge. This tool is currently not the most efficient depression-removal algorithm available in WhiteboxTools; fill_depressions and breach_depressions_least_cost are both more efficient and often produce better, lower-impact results.

The user may optionally specify the size of the elevation increment used to solve flats (flat_increment), although it is best not to specify this optional value and to let the algorithm determine the most suitable value itself.

Reference

Planchon, O. and Darboux, F., 2002. A fast, simple and versatile algorithm to fill the depressions of digital elevation models. Catena, 46(2-3), pp.159-176.

See Also

fill_depressions, breach_depressions_least_cost

Function Signature

def fill_depressions_planchon_and_darboux(self, dem: Raster, fix_flats: bool = True, flat_increment: float = float('nan')) -> Raster: ...

fill_depressions_wang_and_liu

This tool can be used to fill all of the depressions in a digital elevation model (DEM) and to remove the flat areas. This is a common pre-processing step required by many flow-path analysis tools to ensure continuous flow from each grid cell to an outlet located along the grid edge. The fill_depressions_wang_and_liu algorithm is based on the computationally efficient approach of examining each cell based on its spill elevation, starting from the edge cells, and visiting cells from lowest order using a priority queue. As such, it is based on the algorithm first proposed by Wang and Liu (2006). However, it is currently not the most efficient depression-removal algorithm available in WhiteboxTools; fill_depressions and breach_depressions_least_cost are both more efficient and often produce better, lower-impact results.

If the input DEM has gaps, or missing-data holes, that contain NoData values, it is better to use the fill_missing_data tool to repair these gaps. This tool will interpolate values across the gaps and produce a more natural-looking surface than the flat areas that are produced by depression filling. Importantly, the fill_depressions tool algorithm implementation assumes that there are no 'donut hole' NoData gaps within the area of valid data. Any NoData areas along the edge of the grid will simply be ignored and will remain NoData areas in the output image.

The user may optionally specify the size of the elevation increment used to solve flats (flat_increment), although it is best not to specify this optional value and to let the algorithm determine the most suitable value itself.

Reference

Wang, L. and Liu, H. 2006. An efficient method for identifying and filling surface depressions in digital elevation models for hydrologic analysis and modelling. International Journal of Geographical Information Science, 20(2): 193-213.

See Also

fill_depressions, breach_depressions_least_cost, breach_depressions_least_cost, fill_missing_data

Function Signature

def fill_depressions_wang_and_liu(self, dem: Raster, fix_flats: bool = True, flat_increment: float = float('nan')) -> Raster: ...

fill_missing_data

This tool can be used to fill in small gaps in a raster or digital elevation model (DEM). The gaps, or holes, must have recognized NoData values. If gaps do not currently have this characteristic, use the set_nodata_value tool and ensure that the data are stored using a raster format that supports NoData values. All valid, non-NoData values in the input raster will be assigned the same value in the output image.

The algorithm uses an inverse-distance weighted (IDW) scheme based on the valid values on the edge of NoData gaps to estimate gap values. The user must specify the filter size (filter), which determines the size of gap that is filled, and the IDW weight (weight).

The filter size, specified in grid cells, is used to determine how far the algorithm will search for valid, non-NoData values. Therefore, setting a larger filter size allows for the filling of larger gaps in the input raster.

The no_edges flag can be used to exclude NoData values that are connected to the edges of the raster. It is usually the case that irregularly shaped DEMs have large regions of NoData values along the containing raster edges. This flag can be used to exclude these regions from the gap-filling operation, leaving only interior gaps for filling.

See Also

set_nodata_value

Function Signature

def fill_missing_data(self, dem: Raster, filter_size: int = 11, weight: float = 2.0, exclude_edge_nodata: bool = False) -> Raster: ...

fill_pits

This tool can be used to remove pits from a digital elevation model (DEM). Pits are single grid cells with no downslope neighbours. They are important because they impede overland flow-paths. This tool will remove any pits in the input DEM that can be resolved by raising the elevation of the pit such that flow will continue past the pit cell to one of the downslope neighbours. Notice that this tool can be a useful pre-processing technique before running one of the more robust depression breaching (breach_depressions_least_cost) or filling (fill_depressions) techniques, which are designed to remove larger depression features.

See Also

breach_depressions_least_cost, fill_depressions, breach_single_cell_pits

Function Signature

def fill_pits(self, dem: Raster) -> Raster: ...

filter_lidar_classes

This tool can be used to remove points within a LAS LiDAR file that possess certain specified class values. The user must input the names of the input (input) and output (output) LAS files and the class values to be excluded (exclude_cls). Class values are specified by their numerical values, such that:

Classification ValueMeaning
0Created never classified
1Unclassified
2Ground
3Low Vegetation
4Medium Vegetation
5High Vegetation
6Building
7Low Point (noise)
8Reserved
9Water
10Rail
11Road Surface
12Reserved
13Wire – Guard (Shield)
14Wire – Conductor (Phase)
15Transmission Tower
16Wire-structure Connector (e.g. Insulator)
17Bridge Deck
18High noise

Thus, to filter out low and high noise points from a point cloud, specify exclude_cls='7,18'. Class ranges may also be specified, e.g. exclude_cls='3-5,7,18'. Notice that usage of this tool assumes that the LAS file has underwent a comprehensive point classification, which not all point clouds have had. Use the lidar_info tool determine the distribution of various class values in your file.

See Also

lidar_info

Function Signature

def filter_lidar_classes(self, input: Lidar, exclusion_classes: List[int]) -> Lidar: ...

filter_lidar_scan_angles

Function Signature

def filter_lidar_scan_angles(self, in_lidar: Lidar, threshold: int) -> Lidar: ...

filter_raster_features_by_area

This tool takes an input raster (input) containing integer-labelled features, such as the output of the clump tool, and removes all features that are smaller than a user-specified size (threshold), measured in grid cells. The user must specify the replacement value for removed features using the background parameter, which can be either zero or nodata.

See Also

clump

Function Signature

def filter_raster_features_by_area(self, input: Raster, threshold: int, zero_background: bool = False) -> Raster: ...

find_flightline_edge_points

Function Signature

def find_flightline_edge_points(self, in_lidar: Lidar) -> Lidar: ...

find_lowest_or_highest_points

This tool locates the lowest and/or highest cells in a raster and outputs these locations to a vector points file. The user must specify the name of the input raster (input) and the name of the output vector file (output). The user also has the option (out_type) to locate either the lowest value, highest value, or both values. The output vector's attribute table will contain fields for the points XY coordinates and their values.

See Also

extract_raster_values_at_points

Function Signature

def find_lowest_or_highest_points(self, raster: Raster, output_type: str = "lowest") -> Vector: ...

find_main_stem

This tool can be used to identify the main channel in a stream network. The user must input a D8 pointer (flow direction) raster (d8_pointer), and a streams raster (streams_raster). The pointer raster is used to traverse the stream network and should only be created using the d8_pointer tool. By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools:

...
641281
3202
1684

If the pointer file contains ESRI flow direction values instead, you must set esri_pointer=True parameter must be specified.

The streams raster should have been created using one of the DEM-based stream mapping methods, i.e. contributing area thresholding. Stream grid cells are designated in the streams image as all positive, non-zero values. All non-stream cells will be assigned the NoData value in the output image, unless the user sets zero_background=True.

The algorithm operates by traversing each stream and identifying the longest stream-path draining to each outlet. When a confluence is encountered, the traverse follows the branch with the larger distance-to-head.

See Also

d8_pointer

Function Signature

def find_main_stem(self, d8_pointer: Raster, streams_raster: Raster, esri_pointer: bool = False, zero_background: bool = False) -> Raster: ...

find_noflow_cells

This tool can be used to find cells with undefined flow, i.e. no valid flow direction, based on the D8 flow direction algorithm (d8_pointer). These cells are therefore either at the bottom of a topographic depression or in the interior of a flat area. In a digital elevation model (DEM) that has been pre-processed to remove all depressions and flat areas (breach_depressions_least_cost), this condition will only occur along the edges of the grid, otherwise no-flow grid cells can be situation in the interior. The user must specify the name (dem) of the DEM.

See Also

d8_pointer, breach_depressions_least_cost

Function Signature

def find_noflow_cells(self, dem: Raster) -> Raster: ...

find_parallel_flow

This tool can be used to find cells in a stream network grid that possess parallel flow directions based on an input D8 flow-pointer grid (d8_pointer). Because streams rarely flow in parallel for significant distances, these areas are likely errors resulting from the biased assignment of flow direction based on the D8 method.

See Also

d8_pointer

Function Signature

def find_parallel_flow(self, d8_pntr: Raster, streams: Raster) -> Raster: ...

find_patch_edge_cells

This tool will identify all grid cells situated along the edges of patches or class features within an input raster (input). Edge cells in the output raster (output) will have the patch identifier value assigned in the corresponding grid cell. All non-edge cells will be assigned zero in the output raster. Patches (or classes) are designated by positive, non-zero values in the input image. Zero-valued and NoData-valued grid cells are interpreted as background cells by the tool.

See Also

edge_proportion

Function Signature

def find_patch_edge_cells(self, raster: Raster) -> Raster: ...

find_ridges

This tool can be used to identify ridge cells in a digital elevation model (DEM). Ridge cells are those that have lower neighbours either to the north and south or the east and west. Line thinning can optionally be used to create single-cell wide ridge networks by specifying the line_thin parameter.

Function Signature

def find_ridges(self, dem: Raster, line_thin: bool = True) -> Raster: ...

flatten_lakes

This tool can be used to set the elevations contained in a set of input vector lake polygons (lakes) to a consistent value within an input (dem) digital elevation model (DEM). Lake flattening is a common pre-processing step for DEMs intended for use in hydrological applications. This algorithm determines lake elevation automatically based on the minimum perimeter elevation for each lake polygon. The minimum perimeter elevation is assumed to be the lake outlet elevation and is assigned to the entire interior region of lake polygons, excluding island geometries. Note, this tool will not provide satisfactory results if the input vector polygons contain wide river features rather than true lakes. When this is the case, the tool will lower the entire river to the elevation of its mouth, leading to the creation of an artificial gorge.

See Also

fill_depressions

Function Signature

def flatten_lakes(self, dem: Raster, lakes: Vector) -> Raster: ...

flightline_overlap

This tool can be used to map areas of overlapping flightlines in an input LiDAR (LAS) file (input). The output raster file (output) will contain the number of different flightlines that are contained within each grid cell. The user must specify the desired cell size (resolution). The flightline associated with a LiDAR point is assumed to be contained within the point's Point Source ID property. Thus, the tool essentially counts the number of different Point Source ID values among the points contained within each grid cell. If the Point Source ID property is not set, or has been lost, users may with to apply the RecoverFlightlineInfo tool prior to running flightline_overlap.

It is important to set the resolution parameter appropriately, as setting this value too high will yield the mis-characterization of non-overlap areas, and setting the resolution to low will result in fewer than expected overlap areas. An appropriate resolution size value may require experimentation, however a value that is 2-3 times the nominal point spacing has been previously recommended. The nominal point spacing can be determined using the lidar_info tool.

Note that this tool is intended to be applied to LiDAR tile data containing points that have been merged from multiple overlapping flightlines. It is commonly the case that airborne LiDAR data from each of the flightlines from a survey are merged and then tiled into 1 km2 tiles, which are the target dataset for this tool.

Like many of the LiDAR related tools, the input and output file parameters are optional. If left unspecified, the tool will locate all valid LiDAR files within the current Whitebox working directory and use these for calculation (specifying the output raster file name based on the associated input LiDAR file). This can be a helpful way to run the tool on a batch of user inputs within a specific directory.

See Also

classify_overlap_points, RecoverFlightlineInfo, lidar_info

Function Signature

def flightline_overlap(self, input_lidar: Lidar, resolution: float = 1.0) -> Raster: ...

flip_image

This tool can be used to flip, or reflect, an image (input) either vertically, horizontally, or both. The axis of reflection is specified using the direction parameter. The input image is not reflected in place; rather, the reflected image is stored in a separate output file.

Function Signature

def flip_image(self, raster: Raster, direction: str = "v") -> Raster: ...

flood_order

This tool takes an input digital elevation model (DEM) and creates an output raster where every grid cell contains the flood order of that cell within the DEM. The flood order is the sequence of grid cells that are encountered during a search, starting from the raster grid edges and the lowest grid cell, moving inward at increasing elevations. This is in fact similar to how the highly efficient Wang and Liu (2006) depression filling algorithm and the Breach Depressions (Fast) operates. The output flood order raster contains the sequential order, from lowest edge cell to the highest pixel in the DEM.

Like the fill_depressions tool, flood_order will read the entire DEM into memory. This may make the algorithm ill suited to processing massive DEMs except where the user's computer has substantial memory (RAM) resources.

Reference

Wang, L., and Liu, H. (2006). An efficient method for identifying and filling surface depressions in digital elevation models for hydrologic analysis and modelling. International Journal of Geographical Information Science, 20(2), 193-213.

See Also

fill_depressions

Function Signature

def flood_order(self, dem: Raster) -> Raster: ...

flow_accum_full_workflow

Resolves all of the depressions in a DEM, outputting a breached DEM, an aspect-aligned non-divergent flow pointer, and a flow accumulation raster.

Function Signature

def flow_accum_full_workflow(self, dem: Raster, out_type: str = "sca", log_transform: bool = False, clip: bool = False, esri_pntr: bool = False) -> Tuple[Raster, Raster, Raster]: ...

flow_length_diff

FlowLengthDiff calculates the local maximum absolute difference in downslope flowpath length, which is useful in mapping drainage divides and ridges.

See Also

max_branch_length

Function Signature

def flow_length_diff(self, d8_pointer: Raster, esri_pointer: bool = False, log_transform: bool = False) -> Raster: ...

gamma_correction

This tool performs a gamma colour correction transform on an input image (input), such that each input pixel value (zin) is mapped to the corresponding output value (zout) as:

zout = zingamma

The user must specify the value of the gamma parameter. The input image may be of either a greyscale or RGB colour composite data type.

Function Signature

def gamma_correction(self, raster: Raster, gamma_value: float = 0.5) -> Raster: ...

gaussian_contrast_stretch

This tool performs a Gaussian stretch on a raster image. The observed histogram of the input image is fitted to a Gaussian histogram, i.e. normal distribution. A histogram matching technique is used to map the values from the input image onto the output Gaussian distribution. The user must input the number of tones (num_tones) used.

This tool is related to the more general histogram_matching tool, which can be used to fit any frequency distribution to an input image, and other contrast enhancement tools such as histogram_equalization, min_max_contrast_stretch, percentage_contrast_stretch, sigmoidal_contrast_stretch, and standard_deviation_contrast_stretch.

See Also

PiecewiseContrastStretch, histogram_equalization, min_max_contrast_stretch, percentage_contrast_stretch, sigmoidal_contrast_stretch, standard_deviation_contrast_stretch, histogram_matching

Function Signature

def gaussian_contrast_stretch(self, raster: Raster, num_tones: int = 256) -> Raster: ...

gaussian_curvature

This tool calculates the Gaussian curvature from a digital elevation model (DEM). Gaussian curvature is the product of maximal and minimal curvatures, and retains values in each point of the topographic surface after its bending without breaking, stretching, and compressing (Florinsky, 2017). Gaussian curvature is measured in units of m-2.

The user must input a DEM (dem).The Z conversion factor (zfactor) is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z Conversion Factor. Curvature values are often very small and as such the user may opt to log-transform the output raster (log). Transforming the values applies the equation by Shary et al. (2002):

Θ' = sign(Θ) ln(1 + 10n|Θ|)

where Θ is the parameter value and n is dependent on the grid cell size.

For DEMs in projected coordinate systems, the tool uses the 3rd-order bivariate Taylor polynomial method described by Florinsky (2016). Based on a polynomial fit of the elevations within the 5x5 neighbourhood surrounding each cell, this method is considered more robust against outlier elevations (noise) than other methods. For DEMs in geographic coordinate systems (i.e. angular units), the tool uses the 3x3 polynomial fitting method for equal angle grids also described by Florinsky (2016).

References

Florinsky, I. (2016). Digital terrain analysis in soil science and geology. Academic Press.

Florinsky, I. V. (2017). An illustrated introduction to general geomorphometry. Progress in Physical Geography, 41(6), 723-752.

Shary P. A., Sharaya L. S. and Mitusov A. V. (2002) Fundamental quantitative methods of land surface analysis. Geoderma 107: 1–32.

See Also

tangential_curvature, profile_curvature, plan_curvature, mean_curvature, minimal_curvature, maximal_curvature

Function Signature

def gaussian_curvature(self, dem: Raster, log_transform: bool = False, z_factor: float = 1.0) -> Raster: ...

gaussian_filter

This tool can be used to perform a Gaussian filter on a raster image. A Gaussian filter can be used to emphasize the longer-range variability in an image, effectively acting to smooth the image. This can be useful for reducing the noise in an image. The algorithm operates by convolving a kernel of weights with each grid cell and its neighbours in an image. The weights of the convolution kernel are determined by the 2-dimensional Gaussian (i.e. normal) curve, which gives stronger weighting to cells nearer the kernel centre. It is this characteristic that makes the Gaussian filter an attractive alternative for image smoothing and noise reduction than the mean_filter. The size of the filter is determined by setting the standard deviation parameter (sigma), which is in units of grid cells; the larger the standard deviation the larger the resulting filter kernel. The standard deviation can be any number in the range 0.5-20.

gaussian_filter works with both greyscale and red-green-blue (RGB) colour images. RGB images are decomposed into intensity-hue-saturation (IHS) and the filter is applied to the intensity channel. NoData values in the input image are ignored during processing.

Like many low-pass filters, Gaussian filtering can significantly blur well-defined edges in the input image. The edge_preserving_mean_filter and bilateral_filter offer more robust feature preservation during image smoothing. gaussian_filter is relatively slow compared to the fast_almost_gaussian_filter tool, which offers a fast-running approximatation to a Gaussian filter for larger kernel sizes.

See Also

fast_almost_gaussian_filter, mean_filter, median_filter, rgb_to_ihs

Function Signature

def gaussian_filter(self, raster: Raster, sigma: float = 0.75) -> Raster: ...

geomorphons

This tool can be used to perform a geomorphons landform classification based on an input digital elevation model (dem). The geomorphons concept is based on line-of-sight analysis for the eight topographic profiles in the cardinal directions surrounding each grid cell in the input DEM. The relative sizes of the zenith angle of a profile's maximum elevation angle (i.e. horizon angle) and the nadir angle of a profile's minimum elevation angle are then used to generate a ternary (base-3) digit: 0 when the nadir angle is less than the zenith angle, 1 when the two angles differ by less than a user-defined flatness threshold (threshold), and 2 when the nadir angle is greater than the zenith angle. A ternary number is then derived from the digits assigned to each of the eight profiles, with digits sequenced counter-clockwise from east. This ternary number forms the geomorphons code assigned to the grid cell. There are 38 = 6561 possible codes, although many of these codes are equivalent geomorphons through rotations and reflections. Some of the remaining geomorphons also rarely if ever occur in natural topography. Jasiewicz et al. (2013) identified 10 common landform types by reclassifying related geomorphons codes. The user may choose to output these common forms (forms) rather than the the raw ternary code. These landforms include:

ValueLandform Type
1Flat
2Peak (summit)
3Ridge
4Shoulder
5Spur (convex)
6Slope
7Hollow (concave)
8Footslope
9Valley
10Pit (depression)

One of the main advantages of the geomrophons method is that, being based on minimum/maximum elevation angles, the scale used to estimate the landform type at a site adapts to the surrounding terrain. In principle, choosing a large value of search distance (search) should result in identification of a landform element regardless of its scale.

An experimental feature has been added to correct for global inclination. Global inclination biases the flatness threshold angle becasue it is measured relative to the z-axis, especially in locally flat areas. Including the residuals flag "flattens" the input by converting elevation to residuals of a 2-d linear model.

Reference

Jasiewicz, J., and Stepinski, T. F. (2013). Geomorphons — a pattern recognition approach to classification and mapping of landforms. Geomorphology, 182, 147-156.

See Also

PennockLandformClass

Function Signature

def geomorphons(self, dem: Raster, search_distance: int = 1, flatness_threshold: float = 1.0, flatness_distance: int = 0, skip_distance: int = 0, output_forms: bool = True, analyze_residuals: bool = False) -> Raster: ...

hack_stream_order

This tool can be used to assign the Hack stream order to each link in a stream network. According to this common stream numbering system, the main stream is assigned an order of one. All tributaries to the main stream (i.e. the trunk) are assigned an order of two; tributaries to second-order links are assigned an order of three, and so on. The trunk or main stream of the stream network can be defined either based on the furthest upstream distance, at each bifurcation (i.e. network junction).

Stream order is often used in hydro-geomorphic and ecological studies to quantify the relative size and importance of a stream segment to the overall river system. Unlike some other stream ordering systems, e.g. Horton-Strahler stream order (strahler_stream_order) and Shreve's stream magnitude (shreve_stream_magnitude), Hack's stream ordering method increases from the catchment outlet towards the channel heads. This has the main advantage that the catchment outlet is likely to be accurately located while the channel network extent may be less accurately mapped.

The user must input a streams raster image (streams_raster) and D8 pointer image (d8_pntr). Stream cells are designated in the streams image as all positive, nonzero values. Thus all non-stream or background grid cells are commonly assigned either zeros or NoData values. The pointer image is used to traverse the stream network and should only be created using the D8 algorithm (d8_pointer). Background cells will be assigned the NoData value in the output image, unless the zero_background=True, in which case non-stream cells will be assigned zero values in the output.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, the user should specify esri_pntr=True.

Reference

Hack, J. T. (1957). Studies of longitudinal stream profiles in Virginia and Maryland (Vol. 294). US Government Printing Office.

See Also

horton_stream_order, strahler_stream_order, shreve_stream_magnitude, topological_stream_order

Function Signature

def hack_stream_order(self, d8_pntr: Raster, streams_raster: Raster, esri_pntr: bool = False, zero_background: bool = False) -> Raster: ...

heat_map

This tool is used to generate a raster heat map, or kernel density estimation surface raster from a set of vector points (input). Heat mapping is a visualization and modelling technique used to create the continuous density surface associated with the occurrences of a point phenomenon. Heat maps can therefore be used to identify point clusters by mapping the concentration of event occurrence. For example, heat maps have been used extensively to map the spatial distributions of crime events (i.e. crime mapping) or disease cases.

By default, the tool maps the density of raw occurrence events, however, the user may optionally specify an associated weights field (weights) from the point file's attribute table. When a weights field is specified, these values are simply multiplied by each of the individual components of the density estimate. Weights must be numeric.

The bandwidth parameter (--bandwidth) determines the radius of the kernel used in calculation of the density surface. There are guidelines that statisticians use in determining an appropriate bandwidth for a particular population and data set, but often this parameter is determined through experimentation. The bandwidth of the kernel is a free parameter which exhibits a strong influence on the resulting estimate.

The user must specify the kernel function type (kernel). Options include 'uniform', 'triangular', 'epanechnikov', 'quartic', 'triweight', 'tricube', 'gaussian', 'cosine', 'logistic', 'sigmoid', and 'silverman'; 'quartic' is the default kernel type. Descriptions of each function can be found at the link above.

The characteristics of the output raster (resolution and extent) are determined by one of two optional parameters, cell_size and base. If the user optionally specifies the output grid cell size parameter (cell_size) then the coordinates of the output raster extent are determined by the input vector (i.e. the bounding box) and the specified cell size determines the number of rows and columns. If the user instead specifies the optional base raster file parameter (base), the output raster's coordinates (i.e. north, south, east, west) and row and column count, and therefore, resolution, will be the same as the base file.

Reference

Geomatics (2017) QGIS Heatmap Using Kernel Density Estimation Explained, online resource: https://www.geodose.com/2017/11/qgis-heatmap-using-kernel-density.html visited 02/06/2022.

Function Signature

def heat_map(self, points: Vector, field_name: str, bandwidth: float = 0.0, cell_size: float = 0.0, base_raster: Raster = None, kernel_function: str = "quartic") -> Raster: ...

height_above_ground

This tool normalizes an input LiDAR point cloud (input) such that point z-values in the output LAS file (output) are converted from elevations to heights above the ground, specifically the height above the nearest ground-classified point. The input LAS file must have ground-classified points, otherwise the tool will return an error. The lidar_tophat_transform tool can be used to perform the normalization if a ground classification is lacking.

See Also

lidar_tophat_transform

Function Signature

def height_above_ground(self, input: Lidar) -> Lidar: ...

hexagonal_grid_from_raster_base

This tool can be used to create a hexagonal vector grid. The extent of the hexagonal grid is based on the extent of an input raster base file (base). The user must also specify the hexagonal cell width (width) and whether the hexagonal orientation (orientation) is horizontal or vertical. To use a vector base image instead of a raster, use the hexagonal_grid_from_vector_base tool.

See Also

hexagonal_grid_from_vector_base

Function Signature

def hexagonal_grid_from_raster_base(self, base: Raster, width: float, orientation: str = "h") -> Vector: ...

hexagonal_grid_from_vector_base

This tool can be used to create a hexagonal vector grid. The extent of the hexagonal grid is based on the extent of an input vector base file (base). The user must also specify the hexagonal cell width (width) and whether the hexagonal orientation (orientation) is horizontal or vertical. To use a raster base image instead of a vector, use the hexagonal_grid_from_raster_base tool.

See Also

hexagonal_grid_from_raster_base

Function Signature

def hexagonal_grid_from_vector_base(self, base: Vector, width: float, orientation: str = "h") -> Vector: ...

high_pass_filter

This tool performs a high-pass filter on a raster image. High-pass filters can be used to emphasize the short-range variability in an image. The algorithm operates essentially by subtracting the value at the grid cell at the centre of the window from the average value in the surrounding neighbourhood (i.e. window.)

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

See Also

high_pass_median_filter, mean_filter

Function Signature

def high_pass_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

high_pass_median_filter

This tool performs a high-pass median filter on a raster image. High-pass filters can be used to emphasize the short-range variability in an image. The algorithm operates essentially by subtracting the value at the grid cell at the centre of the window from the median value in the surrounding neighbourhood (i.e. window.)

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

See Also

high_pass_filter, median_filter

Function Signature

def high_pass_median_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11, sig_digits: int = 2) -> Raster: ...

highest_position

This tool identifies the stack position (index) of the maximum value within a raster stack on a cell-by-cell basis. For example, if five raster images (inputs) are input to the tool, the output raster (output) would show which of the five input rasters contained the highest value for each grid cell. The index value in the output raster is the zero-order number of the raster stack, i.e. if the highest value in the stack is contained in the first image, the output value would be 0; if the highest stack value were the second image, the output value would be 1, and so on. If any of the cell values within the stack is NoData, the output raster will contain the NoData value for the corresponding grid cell. The index value is related to the order of the input images.

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

lowest_position, pick_from_list

Function Signature

def highest_position(self, input_rasters: List[Raster]) -> Raster: ...

hillshade

This tool performs a hillshade operation (also called shaded relief) on an input digital elevation model (DEM). The user must input a DEM. Other parameters that must be specified include the illumination source azimuth (azimuth), or sun direction (0-360 degrees), the illumination source altitude (altitude; i.e. the elevation of the sun above the horizon, measured as an angle from 0 to 90 degrees) and the Z conversion factor (zfactor). The Z conversion factor is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z conversion factor. If the DEM is in the geographic coordinate system (latitude and longitude), the following equation is used:

zfactor = 1.0 / (111320.0 x cos(mid_lat))

where mid_lat is the latitude of the centre of the raster, in radians.

The hillshade value (HS) of a DEM grid cell is calculate as:

HS = tan(s) / [1 - tan(s)2]0.5 x [sin(Alt) / tan(s) - cos(Alt) x sin(Az - a)]

where s and a are the local slope gradient and aspect (orientation) respectively and Alt and Az are the illumination source altitude and azimuth respectively. Slope and aspect are calculated using Horn's (1981) 3rd-order finate difference method.

Reference

Gallant, J. C., and J. P. Wilson, 2000, Primary topographic attributes, in Terrain Analysis: Principles and Applications, edited by J. P. Wilson and J. C. Gallant pp. 51-86, John Wiley, Hoboken, N.J.

See Also

hypsometrically_tinted_hillshade, multidirectional_hillshade, aspect, slope

Function Signature

def hillshade(self, dem: Raster, azimuth: float = 315.0, altitude: float = 30.0, z_factor: float = 1.0) -> Raster: ...

hillslopes

This tool decrements (lowers) the elevations of pixels within an input digital elevation model (DEM) (dem) along an input vector stream network (streams) at the sites of road (roads) intersections. In addition to the input data layers, the user must specify the output raster DEM (output), and the maximum road embankment width (width), in map units. The road width parameter is used to determine the length of channel along stream lines, at the junctions between streams and roads, that the burning (i.e. decrementing) operation occurs. The algorithm works by identifying stream-road intersection cells, then traversing along the rasterized stream path in the upstream and downstream directions by half the maximum road embankment width. The minimum elevation in each stream traversal is identified and then elevations that are higher than this value are lowered to the minimum elevation during a second stream traversal.

Reference

Lindsay JB. 2016. The practice of DEM stream burning revisited. Earth Surface Processes and Landforms, 41(5): 658–668. DOI: 10.1002/esp.3888

See Also

raster_streams_to_vector, rasterize_streams

Function Signature

def hillslopes(self, d8_pntr: Raster, streams: Raster, esri_pntr: bool = False) -> Raster: ...

histogram_equalization

This tool alters the cumulative distribution function (CDF) of a raster image to match, as closely as possible, the CDF of a uniform distribution. Histogram equalization works by first calculating the histogram of the input image. This input histogram is then converted into a CDF. Each grid cell value in the input image is then mapped to the corresponding value in the uniform distribution's CDF that has an equivalent (or as close as possible) cumulative probability value. Histogram equalization provides a very effective means of performing image contrast adjustment in an efficient manner with little need for human input.

The user must specify the name of the input image to perform histogram equalization on. The user must also specify the number of tones, corresponding to the number of histogram bins used in the analysis.

histogram_equalization is related to the histogram_matching_two_images tool (used when an image's CDF is to be matched to a reference CDF derived from a reference image). Similarly, histogram_matching, and gaussian_contrast_stretch are similarly related tools frequently used for image contrast adjustment, where the reference CDFs are uniform and Gaussian (normal) respectively.

Notes:

  • The algorithm can introduces gaps in the histograms (steps in the CDF). This is to be expected because the histogram is being distorted. This is more prevalent for integer-level images.
  • Histogram equalization is not appropriate for images containing categorical (class) data.

See Also

PiecewiseContrastStretch, histogram_matching, histogram_matching_two_images, gaussian_contrast_stretch

Function Signature

def histogram_equalization(self, raster: Raster, num_tones: int = 256) -> Raster: ...

histogram_matching

This tool alters the cumulative distribution function (CDF) of a raster image to match, as closely as possible, the CDF of a reference histogram. Histogram matching works by first calculating the histogram of the input image. This input histogram and reference histograms are each then converted into CDFs. Each grid cell value in the input image is then mapped to the corresponding value in the reference CDF that has an equivalent (or as close as possible) cumulative probability value. Histogram matching provides the most flexible means of performing image contrast adjustment.

The reference histogram must be specified to the tool in the form of a text file (.txt), provided using the histo_file flag. This file must contain two columns (delimited by a tab, space, comma, colon, or semicolon) where the first column contains the x value (i.e. the values that will be assigned to the grid cells in the output image) and the second column contains the frequency or probability. Note that 1) the file must not contain a header row, 2) each x value/frequency pair must be on a separate row. It is possible to create this type of histogram using the wide range of distribution tools available in most spreadsheet programs (e.g. Excel or LibreOffice's Calc program). You must save the file as a text-only (ASCII) file.

histogram_matching is related to the histogram_matching_two_images tool, which can be used when a reference CDF can be derived from a reference image. histogram_equalization and gaussian_contrast_stretch are similarly related tools frequently used for image contrast adjustment, where the reference CDFs are uniform and Gaussian (normal) respectively.

Notes:

  • The algorithm can introduces gaps in the histograms (steps in the CDF). This is to be expected because the histogram is being distorted. This is more prevalent for integer-level images.
  • Histogram matching is not appropriate for images containing categorical (class) data.
  • This tool is not intended for images containing RGB data. If this is the case, the colour channels should be split using the split_colour_composite tool.

See Also

histogram_matching_two_images, histogram_equalization, gaussian_contrast_stretch, split_colour_composite

Function Signature

def histogram_matching(self, image: Raster, histogram: List[List[float]], histo_is_cumulative: bool = False) -> Raster: ...

histogram_matching_two_images

This tool alters the cumulative distribution function (CDF) of a raster image to match, as closely as possible, the CDF of a reference image. Histogram matching works by first calculating the histograms of the input image (i.e. the image to be adjusted) and the reference image. These histograms are then converted into CDFs. Each grid cell value in the input image is then mapped to the corresponding value in the reference CDF that has the an equivalent (or as close as possible) cumulative probability value. A common application of this is to match the images from two sensors with slightly different responses, or images from the same sensor, but the sensor's response is known to change over time.The size of the two images (rows and columns) do not need to be the same, nor do they need to be geographically overlapping.

histogram_matching_two_images is related to the histogram_matching tool, which can be used when a reference CDF is used directly rather than deriving it from a reference image. histogram_equalization and gaussian_contrast_stretch are similarly related tools, where the reference CDFs are uniform and Gaussian (normal) respectively.

The algorithm may introduces gaps in the histograms (steps in the CDF). This is to be expected because the histograms are being distorted. This is more prevalent for integer-level images. Histogram matching is not appropriate for images containing categorical (class) data. It is also not intended for images containing RGB data, in which case, the colour channels should be split using the split_colour_composite tool.

See Also

histogram_matching, histogram_equalization, gaussian_contrast_stretch, split_colour_composite

Function Signature

def histogram_matching_two_images(self, image1: Raster, image2: Raster) -> Raster: ...

hole_proportion

This calculates the proportion of the total area of a polygon's holes (i.e. islands) relative to the area of the polygon's hull. It can be a useful measure of shape complexity, or how discontinuous a patch is. The user must specify the name of the input vector file and the output data will be contained within the input vector's database file as a new field (HOLE_PROP).

See Also

ShapeComplexityIndex, elongation_ratio, perimeter_area_ratio

Function Signature

def hole_proportion(self, input: Vector) -> Vector: ...

horizon_angle

This tool calculates the horizon angle (Sx), i.e. the maximum slope along a specified azimuth (0-360 degrees) for each grid cell in an input digital elevation model (DEM). Horizon angle is sometime referred to as the maximum upwind slope in wind exposure/sheltering studies. Positive values can be considered sheltered with respect to the azimuth and negative values are exposed. Thus, Sx is a measure of exposure to a wind from a specific direction. The algorithm works by tracing a ray from each grid cell in the direction of interest and evaluating the slope for each location in which the DEM grid is intersected by the ray. Linear interpolation is used to estimate the elevation of the surface where a ray does not intersect the DEM grid precisely at one of its nodes.

The user is able to constrain the maximum search distance (max_dist) for the ray tracing by entering a valid maximum search distance value (in the same units as the X-Y coordinates of the input raster DEM). If the maximum search distance is left blank, each ray will be traced to the edge of the DEM, which will add to the computational time.

Maximum upwind slope should not be calculated for very extensive areas over which the Earth's curvature must be taken into account. Also, this index does not take into account the deflection of wind by topography. However, averaging the horizon angle over a window of directions can yield a more robust measure of exposure, compensating for the deflection of wind from its regional average by the topography. For example, if you are interested in measuring the exposure of a landscape to a northerly wind, you could perform the following calculation:

Sx(N) = [Sx(345)+Sx(350)+Sx(355)+Sx(0)+Sx(5)+Sx(10)+Sx(15)] / 7.0

Ray-tracing is a highly computationally intensive task and therefore this tool may take considerable time to operate for larger sized DEMs. Maximum upwind slope is best displayed using a Grey scale palette that is inverted.

Horizon angle is best visualized using a white-to-black palette and rescaled from approximately -10 to 70 (see below for an example of horizon angle calculated at a 150-degree azimuth).

See Also

time_in_daylight

Function Signature

def horizon_angle(self, dem: Raster, azimuth: float = 0.0, max_dist: float = float('inf')) -> Raster: ...

horton_stream_order

This tool can be used to assign the Horton stream order to each link in a stream network. Stream ordering is often used in hydro-geomorphic and ecological studies to quantify the relative size and importance of a stream segment to the overall river system. There are several competing stream ordering schemes. Based on to this common stream numbering system, headwater stream links are assigned an order of one. Stream order only increases downstream when two links of equal order join, otherwise the downstream link is assigned the larger of the two link orders.

Strahler order and Horton order are similar approaches to assigning stream network hierarchy. Horton stream order essentially starts with the Strahler order scheme, but subsequently replaces each of the assigned stream order value along the main trunk of the network with the order value of the outlet. The main channel is not treated differently compared with other tributaries in the Strahler ordering scheme.

The user must specify input a streams raster image (streams_raster) and D8 pointer image (d8_pntr). Stream cells are designated in the streams image as all positive, nonzero values. Thus all non-stream or background grid cells are commonly assigned either zeros or NoData values. The pointer image is used to traverse the stream network and should only be created using the D8 algorithm (d8_pointer). Background cells will be assigned the NoData value in the output image, unless the user specifies zero_background=True, in which case non-stream cells will be assigned zero values in the output.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, the user must set esri_pntr=True.

Reference Horton, R. E. (1945). Erosional development of streams and their

drainage basins; hydrophysical approach to quantitative morphology. Geological society of America bulletin, 56(3), 275-370.

See Also

hack_stream_order, shreve_stream_magnitude, strahler_stream_order, topological_stream_order

Function Signature

def horton_stream_order(self, d8_pntr: Raster, streams_raster: Raster, esri_pntr: bool = False, zero_background: bool = False) -> Raster: ...

hypsometric_analysis

This tool can be used to derive the hypsometric curve, or area-altitude curve, of one or more input digital elevation models (DEMs) ('inputs'). A hypsometric curve is a histogram or cumulative distribution function of elevations in a geographical area.

See Also

SlopeVsElevationPlot

Function Signature

def hypsometric_analysis(self, dem_rasters: List[Raster], output_html_file: str, watershed_rasters: List[Raster] = None) -> None: ...

hypsometrically_tinted_hillshade

This tool creates a hypsometrically tinted shaded relief (Swiss hillshading) image from an input digital elevation model (DEM). The tool combines a colourized version of the DEM with varying illumination provided by a hillshade image, to produce a composite relief model that can be used to visual topography for more effective interpretation of landscapes. The output of the tool is a 24-bit red-green-blue (RGB) colour image.

The user must input a DEM. Other parameters that must be specified include the illumination source azimuth (azimuth), or sun direction (0-360 degrees), the illumination source altitude (altitude; i.e. the elevation of the sun above the horizon, measured as an angle from 0 to 90 degrees), the hillshade weight (hs_weight; 0-1), image brightness (brightness; 0-1), and atmospheric effects (atmospheric; 0-1). The hillshade weight can be used to increase or subdue the relative prevalence of the hillshading effect in the output image. The image brightness parameter is used to create an overall brighter or darker version of the terrain rendering; note however, that very high values may over-saturate the well-illuminated portions of the terrain. The atmospheric effects parameter can be used to introduce a haze or atmosphere effect to the output image. It is intended to reproduce the effect of viewing mountain valley bottoms through a thicker and more dense atmosphere. Values greater than zero will introduce a slightly blue tint, particularly at lower altitudes, blur the hillshade edges slightly, and create a random haze-like speckle in lower areas. The user must also specify the Z conversion factor (zfactor). The Z conversion factor is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z conversion factor. If the DEM is in the geographic coordinate system (latitude and longitude), the following equation is used:

zfactor = 1.0 / (111320.0 x cos(mid_lat))

where mid_lat is the latitude of the centre of the raster, in radians.

See Also

hillshade, multidirectional_hillshade, aspect, slope

Function Signature

def hypsometrically_tinted_hillshade(self, dem: Raster, solar_altitude: float = 45.0, hillshade_weight: float = 0.5, brightness: float = 0.5, atmospheric_effects: float = 0.0, palette: str = "atlas", reverse_palette: bool = False, full_360_mode: bool = False, z_factor: float = 1.0) -> Raster: ...

idw_interpolation

points or a fixed neighbourhood size. This tool is currently configured to perform the later only, using a FixedRadiusSearch structure. Using a fixed number of neighbours will require use of a KD-tree structure. I've been testing one Rust KD-tree library but its performance does not appear to be satisfactory compared to the FixedRadiusSearch. I will need to explore other options here.

Another change that will need to be implemented is the use of a nodal function. The original Whitebox GAT tool allows for use of a constant or a quadratic. This tool only allows the former.

Function Signature

def idw_interpolation(self, points: Vector, field_name: str = "FID", use_z: bool = False, weight: float = 2.0, radius: float = 0.0, min_points: int = 0, cell_size: float = 0.0, base_raster: Raster = None) -> Raster: ...

ihs_to_rgb

This tool transforms three intensity, hue, and saturation (IHS; sometimes HSI or HIS) raster images into three equivalent multispectral images corresponding with the red, green, and blue channels of an RGB composite. Intensity refers to the brightness of a color, hue is related to the dominant wavelength of light and is perceived as color, and saturation is the purity of the color (Koutsias et al., 2000). There are numerous algorithms for performing a red-green-blue (RGB) to IHS transformation. This tool uses the transformation described by Haydn (1982). Note that, based on this transformation, the input IHS values must follow the ranges:

0 < I < 1

0 < H < 2PI

0 < S < 1

The output red, green, and blue images will have values ranging from 0 to 255. The user must specify the names of the intensity, hue, and saturation images (intensity, hue, saturation). These images will generally be created using the rgb_to_ihs tool. The user must also specify the names of the output red, green, and blue images (red, green, blue). Image enhancements, such as contrast stretching, are often performed on the individual IHS components, which are then inverse transformed back in RGB components using this tool. The output RGB components can then be used to create an improved color composite image.

References

Haydn, R., Dalke, G.W. and Henkel, J. (1982) Application of the IHS color transform to the processing of multisensor data and image enhancement. Proc. of the Inter- national Symposium on Remote Sensing of Arid and Semiarid Lands, Cairo, 599-616.

Koutsias, N., Karteris, M., and Chuvico, E. (2000). The use of intensity-hue-saturation transformation of Landsat-5 Thematic Mapper data for burned land mapping. Photogrammetric Engineering and Remote Sensing, 66(7), 829-840.

See Also

rgb_to_ihs, balance_contrast_enhancement, direct_decorrelation_stretch

Function Signature

def ihs_to_rgb(self, intensity: Raster, hue: Raster, saturation: Raster) -> Tuple[Raster, Raster, Raster]: ...

image_autocorrelation

Spatial autocorrelation describes the extent to which a variable is either dispersed or clustered through space. In the case of a raster image, spatial autocorrelation refers to the similarity in the values of nearby grid cells. This tool measures the spatial autocorrelation of a raster image using the global Moran's I statistic. Moran's I varies from -1 to 1, where I = -1 indicates a dispersed, checkerboard type pattern and I = 1 indicates a clustered (smooth) surface. I = 0 occurs for a random distribution of values. image_autocorrelation computes Moran's I for the first lag only, meaning that it only takes into account the variability among the immediate neighbors of each grid cell.

The user must specify the names of one or more input raster images. In addition, the user must specify the contiguity type (contiguity; Rook's, King's, or Bishop's), which describes which neighboring grid cells are examined for the analysis. The following figure describes the available cases:

Rook's contiguity

...
010
1X1
010

Kings's contiguity

...
111
1X1
111

Bishops's contiguity

...
101
0X0
101

The tool outputs an HTML report (output) which, for each input image (input), reports the Moran's I value and the variance, z-score, and p-value (significance) under normal and randomization sampling assumptions.

Use the image_correlation tool instead when there is need to determine the correlation among multiple raster inputs.

**NoData **values in the input image are ignored during the analysis.

See Also

image_correlation, image_correlation_neighbourhood_analysis

Function Signature

def image_autocorrelation(self, rasters: List[Raster], output_html_file: str, contiguity_type: str = "bishop") -> None: ...

image_correlation

This tool can be used to estimate the Pearson product-moment correlation coefficient (r) between two or more input images (inputs). The r-value is a measure of the linear association in the variation of the input variables (images, in this case). The coefficient ranges from -1.0, indicated a perfect negative linear association, to 1.0, indicated a perfect positive linear association. An r-value of 0.0 indicates no correlation between the test variables.

Note that this index is a measure of the linear association; two variables may be strongly related by a non-linear association (e.g. a power function curve) which will lead to an apparent weak association based on the Pearson coefficient. In fact, non-linear associations are very common among spatial variables, e.g. terrain indices such as slope and contributing area. In such cases, it is advisable that the input images are transformed prior to the estimation of the Pearson coefficient, or that an alternative, non-parametric statistic be used, e.g. the Spearman rank correlation coefficient.

The user must specify the names of two or more input images (inputs). All input images must share the same grid, as the coefficient requires a comparison of a pair of images on a grid-cell-by-grid-cell basis. If more than two image names are selected, the correlation coefficient will be calculated for each pair of images and reported in the HTML output report (output) as a correlation matrix. Caution must be exercised when attempted to estimate the significance of a correlation coefficient derived from image data. The very high N-value (essentially the number of pixels in the image pair) means that even small correlation coefficients can be found to be statistically significant, despite being practically insignificant.

NoData values in either of the two input images are ignored during the calculation of the correlation between images.

See Also

image_correlation_neighbourhood_analysis, image_regression, image_autocorrelation

Function Signature

def image_correlation(self, rasters: List[Raster], output_html_file: str) -> None: ...

image_correlation_neighbourhood_analysis

This tool can be used to perform nieghbourhood-based (i.e. using roving search windows applied to each grid cell) correlation analysis on two input rasters (input1 and input2). The tool outputs a correlation value raster (output1) and a significance (p-value) raster (output2). Additionally, the user must specify the size of the search window (filter) and the correlation statistic (stat). Options for the correlation statistic include pearson, kendall, and spearman. Notice that Pearson's r is the most computationally efficient of the three correlation metrics but is unsuitable when the input distributions are non-linearly associated, in which case, either Spearman's Rho or Kendall's tau-b correlations are more suited. Both Spearman and Kendall correlations evaluate monotonic associations without assuming linearity in the relation. Kendall's tau-b is by far the most computationally expensive of the three statistics and may not be suitable to larger sized search windows.

See Also

image_correlation, image_regression

Function Signature

def image_correlation_neighbourhood_analysis(self, raster1: Raster, raster2: Raster, filter_size: int = 11, correlation_stat: str = "pearson") -> Tuple[Raster, Raster]: ...

image_regression

This tool performs a bivariate linear regression analysis on two input raster images. The first image (i1) is considered to be the independent variable while the second image (i2) is considered to be the dependent variable in the analysis. Both input images must share the same grid, as the coefficient requires a comparison of a pair of images on a grid-cell-by-grid-cell basis. The tool will output an HTML report (output) summarizing the regression model, an Analysis of Variance (ANOVA), and the significance of the regression coefficients. The regression residuals can optionally be output as a new raster image (out_residuals) and the user can also optionally specify to standardize the residuals (standardize).

Note that the analysis performs a linear regression; two variables may be strongly related by a non-linear association (e.g. a power function curve) which will lead to an apparently weak fitting regression model. In fact, non-linear relations are very common among spatial variables, e.g. terrain indices such as slope and contributing area. In such cases, it is advisable that the input images are transformed prior to the analysis.

NoData values in either of the two input images are ignored during the calculation of the correlation between images.

Example usage

import whitebox_workflow

See Also

image_correlation, image_correlation_neighbourhood_analysis

Function Signature

def image_regression(self, independent_variable: Raster, dependent_variable: Raster, output_html_file: str, standardize_residuals: bool = False, output_scattergram: bool = False, num_samples: int = 1000) -> Raster: ...

image_stack_profile

This tool can be used to plot an image stack profile (i.e. a signature) for a set of points (points) and a multispectral image stack (inputs). The tool outputs an interactive SVG line graph embedded in an HTML document. If the input points vector contains multiple points, each input point will be associated with a single line in the output plot. The order of vertices in each signature line is determined by the order of images specified in the inputs parameter. At least two input images are required to run this operation. Note that this tool does not require multispectral images as inputs; other types of data may also be used as the image stack. Also note that the input images should be single-band, continuous greytone rasters. RGB colour images are not good candidates for this tool.

If you require the raster values to be saved in the vector points file's attribute table, or if you need the raster values to be output as text, you may use the extract_raster_values_at_points tool instead.

See Also

extract_raster_values_at_points

Function Signature

def image_stack_profile(self, images: List[Raster], points: Vector, output_html_file: str) -> None: ...

impoundment_size_index

This tool can be used to calculate the impoundment size index (ISI) from a digital elevation model (DEM). The ISI is a land-surface parameter related to the size of the impoundment that would result from inserting a dam of a user-specified maximum length (damlength) into each DEM grid cell. The tool requires the user to specify the name of one or more of the possible outputs, which include the mean flooded depth (out_mean), the maximum flooded depth (out_max), the flooded volume (out_volume), the flooded area (out_area), and the dam height (out_dam_height).

Please note that this tool performs an extremely complex and computationally intensive flow-accumulation operation. As such, it may take a substantial amount of processing time and may encounter issues (including memory issues) when applied to very large DEMs. It is not necessary to pre-process the input DEM (dem) to remove topographic depressions and flat areas. The internal flow-accumulation operation will not be confounded by the presence of these features.

Reference

Lindsay, JB (2015) Modelling the spatial pattern of potential impoundment size from DEMs. Online resource: Whitebox Blog

See Also

insert_dams, stochastic_depression_analysis

Function Signature

def impoundment_size_index(self, dem: Raster, max_dam_length: float, output_mean: bool = False, output_max: bool = False, output_volume: bool = False, output_area: bool = False, output_height: bool = False) -> Tuple[Union[Raster, None], Union[Raster, None], Union[Raster, None], Union[Raster, None], Union[Raster, None]]: ...

insert_dams

This tool can be used to insert dams at one or more user-specified points (dam_pts), and of a maximum length (damlength), within an input digital elevation model (DEM) (dem). This tool can be thought of as providing the impoundment feature that is calculated internally during a run of the the impoundment size index (ISI) tool for a set of points of interest. from a (DEM).

Reference

Lindsay, JB (2015) Modelling the spatial pattern of potential impoundment size from DEMs. Online resource: Whitebox Blog

See Also

impoundment_size_index, stochastic_depression_analysis

Function Signature

def insert_dams(self, dem: Raster, dam_points: Vector, dam_length: float) -> Raster: ...

integral_image_transform

This tool transforms an input raster image into an integral image, or summed area table. Integral images are the two-dimensional equivalent to a cumulative distribution function. Each pixel contains the sum of all pixels contained within the enclosing rectangle above and to the left of a pixel. Images with a very large number of grid cells will likely experience numerical overflow errors when converted to an integral image. Integral images are used in a wide variety of computer vision and digital image processing applications, including texture mapping. They allow for the efficient calculation of very large filters and are the basis of several of WhiteboxTools's image filters.

Reference

Crow, F. C. (1984, January). Summed-area tables for texture mapping. In ACM SIGGRAPH computer graphics (Vol. 18, No. 3, pp. 207-212). ACM.

Function Signature

def integral_image_transform(self, raster: Raster) -> Raster: ...

intersect

The result of the intersect vector overlay operation includes all the feature parts that occur in both input layers, excluding all other parts. It is analogous to the OR logical operator and multiplication in arithmetic. This tool is one of the common vector overlay operations in GIS. The user must specify the names of the input and overlay vector files as well as the output vector file name. The tool operates on vector points, lines, or polygon, but both the input and overlay files must contain the same VectorGeometryType.

The intersect tool is similar to the clip tool. The difference is that the overlay vector layer in a clip operation must always be polygons, regardless of whether the input layer consists of points or polylines.

The attributes of the two input vectors will be merged in the output attribute table. Note, duplicate fields should not exist between the inputs layers, as they will share a single attribute in the output (assigned from the first layer). Multipoint VectorGeometryTypes will simply contain a single output feature identifier (FID) attribute. Also, note that depending on the VectorGeometryType (polylines and polygons), Measure and Z ShapeDimension data will not be transferred to the output geometries. If the input attribute table contains fields that measure the geometric properties of their associated features (e.g. length or area), these fields will not be updated to reflect changes in geometry shape and size resulting from the overlay operation.

See Also

difference, union, symmetrical_difference, clip, erase

Function Signature

def intersect(self, input: Vector, overlay: Vector, snap_tolerance: float = 2.220446049250313e-16) -> Vector: ...

isobasins

This tool can be used to divide a landscape into a group of nearly equal-sized watersheds, known as isobasins. The user must specify the name (dem) of a digital elevation model (DEM), the output raster name (output), and the isobasin target area (size) specified in units of grid cells. The DEM must have been hydrologically corrected to remove all spurious depressions and flat areas. DEM pre-processing is usually achieved using either the breach_depressions_least_cost or fill_depressions tool. Several temporary rasters are created during the execution and stored in memory of this tool.

The tool can optionally (connections) output a CSV table that contains the upstream/downstream connections among isobasins. That is, this table will identify the downstream basin of each isobasin, or will list N/A in the event that there is no downstream basin, i.e. if it drains to an edge. Additionally, the CSV file will contain information about the number of grid cells in each isobasin and the isobasin outlet's row and column number and
flow direction. The output CSV file will have the same name as the output raster, but with a *.csv file extension.

See Also

watershed, basins, breach_depressions_least_cost, fill_depressions

Function Signature

def isobasins(self, dem: Raster, target_size: float, connections: bool = False, csv_file: str = "" ) -> Raster: ...

jenson_snap_pour_points

This tool measures the depth that each grid cell in an input (dem) raster digital elevation model (DEM) lies within a sink feature, i.e. a closed topographic depression. A sink, or depression, is a bowl-like landscape feature, which is characterized by interior drainage and groundwater recharge. The depth_in_sink tool operates by differencing a filled DEM, using the same depression filling method as fill_depressions, and the original surface model.

In addition to the names of the input DEM (dem) and the output raster (output), the user must specify whether the background value (i.e. the value assigned to grid cells that are not contained within sinks) should be set to 0.0 (zero_background) Without this optional parameter specified, the tool will use the NoData value as the background value.

Reference

Antonić, O., Hatic, D., & Pernar, R. (2001). DEM-based depth in sink as an environmental estimator. Ecological Modelling, 138(1-3), 247-254.

See Also

fill_depressions

Function Signature

def jenson_snap_pour_points(self, pour_pts: Vector, streams: Raster, snap_dist: float = 0.0) -> Vector: ...

join_tables

This tool can be used to join (i.e. merge) a vector's attribute table with a second table. The user must specify the name of the vector file (and associated attribute file) as well as the primary key within the table. The primary key (pkey flag) is the field within the table that is being appended to that serves as the identifier. Additionally, the user must specify the name of a second vector from which the data appended into the first table will be derived. The foreign key (fkey flag), the identifying field within the second table that corresponds with the data contained within the primary key in the table, must be specified. Both the primary and foreign keys should either be strings (text) or integer values. Fields containing decimal values are not good candidates for keys. Lastly, the names of the field within the second file to include in the merge operation can also be input (import_field). If the import_field field is not input, all fields in the attribute table of the second file, that are not the foreign key nor FID, will be imported to the first table.

Merging works for one-to-one and many-to-one database relations. A one-to-one relations exists when each record in the attribute table corresponds to one record in the second table and each primary key is unique. Since each record in the attribute table is associated with a geospatial feature in the vector, an example of a one-to-one relation may be where the second file contains AREA and PERIMETER fields for each polygon feature in the vector. This is the most basic type of relation. A many-to-one relation would exist when each record in the first attribute table corresponds to one record in the second file and the primary key is NOT unique. Consider as an example a vector and attribute table associated with a world map of countries. Each country has one or more more polygon features in the shapefile, e.g. Canada has its mainland and many hundred large islands. You may want to append a table containing data about the population and area of each country. In this case, the COUNTRY columns in the attribute table and the second file serve as the primary and foreign keys respectively. While there may be many duplicate primary keys (all of those Canadian polygons) each will correspond to only one foreign key containing the population and area data. This is a many-to-one relation. The join_tables tool does not support one-to-many nor many-to-many relations.

See Also

merge_table_with_csv, reinitialize_attribute_table, export_table_to_csv

Function Signature

def join_tables(self, primary_vector: Vector, primary_key_field: str, foreign_vector: Vector, foreign_key_field: str, import_field: str = "") -> None: ...

k_means_clustering

This tool can be used to perform a k-means clustering operation on two or more input images (inputs), typically several bands of multi-spectral satellite imagery. The tool creates two outputs, including the classified image (output and a classification HTML report (out_html). The user must specify the number of class (classes), which should be known a priori, and the strategy for initializing class clusters (initialize). The initialization strategies include "diagonal" (clusters are initially located randomly along the multi-dimensional diagonal of spectral space) and "random" (clusters are initially located randomly throughout spectral space). The algorithm will continue updating cluster center locations with each iteration of the process until either the user-specified maximum number of iterations (max_iterations) is reached, or until a stability criteria (class_change) is achieved. The stability criteria is the percent of the total number of pixels in the image that are changed among the class values between consecutive iterations. Lastly, the user must specify the minimum allowable number of pixels in a cluster (min_class_size).

Note, each of the input images must have the same number of rows and columns and the same spatial extent because the analysis is performed on a pixel-by-pixel basis. NoData values in any of the input images will result in the removal of the corresponding pixel from the analysis.

See Also

modified_k_means_clustering

Function Signature

def k_means_clustering(self, input_rasters: List[Raster], output_html_file: str = "", num_clusters: int = 5, max_iterations: int = 10, percent_changed_threshold: float = 2.0, initialization_mode: str = "dia", min_class_size: int = 10) -> Raster: ...

k_nearest_mean_filter

This tool performs a k-nearest mean filter on a raster image. A mean filter can be used to emphasize the longer-range variability in an image, effectively acting to smooth or blur the image. This can be useful for reducing the noise in an image. The algorithm operates by calculating the average of a specified number (k) values in a moving window centred on each grid cell. The k values used in the average are those cells in the window with the nearest intensity values to that of the centre cell. As such, this is a type of edge-preserving smoothing filter. The bilateral_filter and edge_preserving_mean_filter are examples of more sophisticated edge-preserving smoothing filters.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

NoData values in the input image are ignored during filtering.

See Also

mean_filter, bilateral_filter, edge_preserving_mean_filter

Function Signature

def k_nearest_mean_filter(self, raster: Raster, filter_size_x: int = 3, filter_size_y: int = 3, k: int = 5) -> Raster: ...

kappa_index

This tool calculates the Kappa index of agreement (KIA), or Cohen's Kappa, for two categorical input raster images (input1 and input2). The KIA is a measure of inter-rater reliability (i.e. classification accuracy) and is widely applied in many fields, notably remote sensing. For example, The KIA is often used as a means of assessing the accuracy of an image classification analysis. The KIA can be interpreted as the percentage improvement that the underlying classification has over and above a random classifier (i.e. random assignment to categories). The user must specify the output HTML file (output). The input images must be of a categorical data type, i.e. contain classes. As a measure of classification accuracy, the KIA is more robust than the overall percent agreement because it takes into account the agreement occurring by chance. A KIA of 0 would indicate that the classifier is no better than random class assignment. In addition to the KIA, this tool will also output the producer's and user's accuracy, the overall accuracy, and the error matrix.

See Also

cross_tabulation

Function Signature

def kappa_index(self, class_raster: Raster, reference_raster: Raster, output_html_file: str = "") -> None: ...

ks_normality_test

This tool will perform a Kolmogorov-Smirnov (K-S) test for normality to evaluate whether the frequency distribution of values within a raster image are drawn from a Gaussian (normal) distribution. The user must specify the name of the raster image. The test can be performed optionally on the entire image or on a random sub-sample of pixel values of a user-specified size. In evaluating the significance of the test, it is important to keep in mind that given a sufficiently large sample, extremely small and non-notable differences can be found to be statistically significant. Furthermore statistical significance says nothing about the practical significance of a difference.

See Also

two_sample_ks_test

Function Signature

def ks_normality_test(self, raster: Raster, output_html_file: str, num_samples: int) -> None: ...

laplacian_filter

This tool can be used to perform a Laplacian filter on a raster image. A Laplacian filter can be used to emphasize the edges in an image. As such, this filter type is commonly used in edge-detection applications. The algorithm operates by convolving a kernel of weights with each grid cell and its neighbours in an image. Four 3x3 sized filters and one 5x5 filter are available for selection. The weights of the kernels are as follows:

3x3(1)

...
0-10
-14-1
0-10

3x3(2)

...
0-10
-15-1
0-10

3x3(3)

...
-1-1-1
-18-1
-1-1-1

3x3(4)

...
1-21
-24-2
1-21

5x5(1)

.....
00-100
0-1-2-10
-1-217-2-1
0-1-2-10
00-100

5x5(2)

.....
00-100
0-1-2-10
-1-216-2-1
0-1-2-10
00-100

The user must specify the variant, including '3x3(1)', '3x3(2)', '3x3(3)', '3x3(4)', '5x5(1)', and '5x5(2)'. The user may also optionally clip the output image distribution tails by a specified amount (e.g. 1%).

See Also

prewitt_filter, sobel_filter

Function Signature

def laplacian_filter(self, raster: Raster, variant: str = "3x3(1)", clip_amount: float = 0.0) -> Raster: ...

laplacian_of_gaussians_filter

The Laplacian-of-Gaussian (LoG) is a spatial filter used for edge enhancement and is closely related to the difference-of-Gaussians filter (DiffOfGaussianFilter). The formulation of the LoG filter algorithm is based on the equation provided in the Hypermedia Image Processing Reference (HIPR) 2. The LoG operator calculates the second spatial derivative of an image. In areas where image intensity is constant, the LoG response will be zero. Near areas of change in intensity the LoG will be positive on the darker side, and negative on the lighter side. This means that at a sharp edge, or boundary, between two regions of uniform but different intensities, the LoG response will be:

  • zero at a long distance from the edge,
  • positive just to one side of the edge,
  • negative just to the other side of the edge,
  • zero at some point in between, on the edge itself.

The user may optionally choose to reflecting the data along image edges. NoData values in the input image are similarly valued in the output. The output raster is of the float data type and continuous data scale.

Reference

Fisher, R. 2004. Hypertext Image Processing Resources 2 (HIPR2). Available online: http://homepages.inf.ed.ac.uk/rbf/HIPR2/roberts.htm

See Also

DiffOfGaussianFilter

Function Signature

def laplacian_of_gaussians_filter(self, raster: Raster, sigma: float = 0.75) -> Raster: ...

las_to_ascii

This tool can be used to convert one or more LAS file, containing LiDAR data, into ASCII files. The user must specify the name(s) of the input LAS file(s) (inputs). Each input file will have a correspondingly named output file with a .csv file extension. CSV files are comma separated value files and contain tabular data with each column corresponding to a field in the table and each row a point value. Fields are separated by commas in the ASCII formatted file. The output point data, each on a separate line, will take the format:

X,Y,Z,INTENSITY,CLASS,RETURN,NUM_RETURN,SCAN_ANGLE

If the LAS file has a point format that contains RGB data, the final three columns will contain the RED, GREEN, and BLUE values respectively. Use the ascii_to_las tool to convert a text file containing LiDAR point data into a LAS file.

See Also

ascii_to_las

Function Signature

def las_to_ascii(self, input_lidar: Lidar) -> None: ...

las_to_shapefile

This tool converts one or more LAS files into a POINT vector. When the input parameter is not specified, the tool grids all LAS files contained within the working directory. The attribute table of the output Shapefile will contain fields for the z-value, intensity, point class, return number, and number of return.

This tool can be used in place of the LasToMultipointShapefile tool when the number of points are relatively low and when the desire is to represent more than simply the x,y,z position of points. Notice however that because each point in the input LAS file will be represented as a separate record in the output Shapefile, the output file will be many time larger than the equivalent output of the LasToMultipointShapefile tool. There is also a practical limit on the total number of records that can be held in a single Shapefile and large LAS files approach this limit. In these cases, the LasToMultipointShapefile tool should be preferred instead.

See Also

LasToMultipointShapefile

Function Signature

def las_to_shapefile(self, input_lidar: Lidar, output_multipoint: bool = False) -> Vector: ...

layer_footprint_raster

This tool creates a vector polygon footprint of the area covered by an input raster grid (input). It will create a vector rectangle corresponding to the bounding box of the input raster.

If input data are irregular shape (i.e. there a boundary of NoData cells) the resulting vector will still correspond to the full grid extent, ignoring the irregular boundary. If this is not the desired effect, you may consider the minimum_bounding_envelope tool instead.

See Also

layer_footprint_vector, minimum_bounding_envelope

Function Signature

def layer_footprint_raster(self, input: Raster) -> Vector: ...

layer_footprint_vector

This tool creates a vector polygon footprint of the area covered by a vector layer. It will create a vector rectangle corresponding to the bounding box. The user must specify the name of the input file (input).

If input data are irregular shape the resulting vector will still correspond to the full grid extent, ignoring the irregular boundary. If this is not the desired effect, you should use the minimum_bounding_envelope tool instead.

See Also

layer_footprint_raster, minimum_bounding_envelope

Function Signature

def layer_footprint_vector(self, input: Vector) -> Vector: ...

lee_filter

The Lee Sigma filter is a low-pass filter used to smooth the input image (input). The user must specify the dimensions of the filter (filterx and filtery) as well as the sigma (sigma) and M (m) parameter.

Reference

Lee, J. S. (1983). Digital image smoothing and the sigma filter. Computer vision, graphics, and image processing, 24(2), 255-269.

See Also

mean_filter, gaussian_filter

Function Signature

def lee_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11, sigma: float = 10.0, m_value: float = 5.0) -> Raster: ...

length_of_upstream_channels

This tool calculates, for each stream grid cell in an input streams raster (streams_raster) the total length of channels upstream. The user must specify the name of a raster containing streams data (streams_raster), where stream grid cells are denoted by all positive non-zero values, and a D8 flow pointer (i.e. flow direction) raster (d8_pointer). The pointer image is used to traverse the stream network and must only be created using the D8 algorithm. Stream cells are designated in the streams image as all values greater than zero. Thus, all non-stream or background grid cells are commonly assigned either zeros or NoData values. Background cells will be assigned the NoData value in the output image, unless the user specifies zero_background=True, in which case non-stream cells will be assigned zero values in the output.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, set esri_pntr=True.

See Also

farthest_channel_head, find_main_stem

Function Signature

def length_of_upstream_channels(self, d8_pointer: Raster, streams_raster: Raster, esri_pointer: bool = False, zero_background: bool = False) -> Raster: ...

lidar_block_maximum

Function Signature

def lidar_block_maximum(self, input_lidar: Lidar, cell_size: float = 1.0) -> Raster: ...

lidar_block_minimum

Function Signature

def lidar_block_minimum(self, input_lidar: Lidar, cell_size: float = 1.0) -> Raster: ...

lidar_classify_subset

This tool classifies points within a user-specified LiDAR point cloud (base) that correspond with points in a subset cloud (subset). The subset point cloud may have been derived by filtering the original point cloud. The user must specify the names of the two input LAS files (i.e. the full and subset clouds) and the class value (subset_class) to assign the matching points. This class value will be assigned to points in the base cloud, overwriting their input class values in the output LAS file (output). Class values should be numerical (integer valued) and should follow the LAS specifications below:

Classification ValueMeaning
0Created never classified
1Unclassified
2Ground
3Low Vegetation
4Medium Vegetation
5High Vegetation
6Building
7Low Point (noise)
8Reserved
9Water
10Rail
11Road Surface
12Reserved
13Wire – Guard (Shield)
14Wire – Conductor (Phase)
15Transmission Tower
16Wire-structure Connector (e.g. Insulator)
17Bridge Deck
18High noise

The user may optionally specify a class value to be assigned to non-subset (i.e. non-matching) points (nonsubset_class) in the base file. If this parameter is not specified, output non-sutset points will have the same class value as the base file.

Function Signature

def lidar_classify_subset(self, base_lidar: Lidar, subset_lidar: Lidar, subset_class_value: int, nonsubset_class_value: int) -> Lidar: ...

lidar_colourize

This tool can be used to add red-green-blue (RGB) colour values to the points contained within an input LAS file (in_lidar), based on the pixel values of an overlapping input colour image (in_image). Ideally, the image has been acquired at the same time as the LiDAR point cloud. If this is not the case, one may expect that transient objects (e.g. cars) in both input data sets will be incorrectly coloured. The input image should overlap in extent with the LiDAR data set and the two data sets should share the same projection. You may use the lidar_tile_footprint tool to determine the spatial extent of the LAS file.

See Also

ColourizeBasedOnClass, ColourizeBasedOnPointReturns, lidar_tile_footprint

Function Signature

def lidar_colourize(self, in_lidar: Lidar, in_image: Raster) -> Lidar: ...

lidar_construct_vector_tin

This tool creates a vector triangular irregular network (TIN) for a set of LiDAR points (input) using a 2D Delaunay triangulation algorithm. LiDAR points may be excluded from the triangulation operation based on a number of criteria, include the point return number (returns), point classification value (exclude_cls), or a minimum (minz) or maximum (maxz) elevation.

For vector points, use the construct_vector_tin tool instead.

See Also

construct_vector_tin

Function Signature

def lidar_construct_vector_tin(self, input_lidar: Lidar, returns_included: str = "all", excluded_classes: List[int] = None, min_elev: float = float('-inf'), max_elev: float = float('inf'), max_triangle_edge_length: float = float('inf')) -> Vector: ...

lidar_digital_surface_model

This tool creates a digital surface model (DSM) from a LiDAR point cloud. A DSM reflects the elevation of the tops of all off-terrain objects (i.e. non-ground features) contained within the data set. For example, a DSM will model the canopy top as well as building roofs. This is in stark contrast to a bare-earth digital elevation model (DEM), which models the ground surface without off-terrain objects present. Bare-earth DEMs can be derived from LiDAR data by interpolating last-return points using one of the other LiDAR interpolators (e.g. lidar_tin_gridding). The algorithm used for interpolation in this tool is based on gridding a triangulation (TIN) fit to top-level points in the input LiDAR point cloud. All points in the input LiDAR data set that are below other neighbouring points, within a specified search radius (radius), and that have a large inter-point slope, are filtered out. Thus, this tool will remove the ground surface beneath as well as any intermediate points within a forest canopy, leaving only the canopy top surface to be interpolated. Similarly, building wall points and any ground points beneath roof overhangs will also be remove prior to interpolation. Note that because the ground points beneath overhead wires and utility lines are filtered out by this operation, these features tend to be appear as 'walls' in the output DSM. If these points are classified in the input LiDAR file, you may wish to filter them out before using this tool (filter_lidar_classes).

The following images show the differences between creating a DSM using the lidar_digital_surface_model and by interpolating first-return points only using the lidar_tin_gridding tool respectively. Note, the images show time_in_daylight, which is a more effective way of hillshading DSMs than the traditional hillshade method. Compare how the DSM created lidar_digital_surface_model tool (above) has far less variability in areas of tree-cover, more effectively capturing the canopy top. As well, notice how building rooftops are more extensive and straighter in the lidar_digital_surface_model DSM image. This is because this method eliminates ground returns beneath roof overhangs before the triangulation operation.

The user must specify the grid resolution of the output raster (resolution), and optionally, the name of the input LiDAR file (input) and output raster (output). Note that if an input LiDAR file (input) is not specified by the user, the tool will search for all valid LiDAR (*.las, *.laz, *.zlidar) files contained within the current working directory. This feature can be very useful when you need to interpolate a DSM for a large number of LiDAR files. Not only does this batch processing mode enable the tool to run in a more optimized parallel manner, but it will also allow the tool to include a small buffer of points extending into adjacent tiles when interpolating an individual file. This can significantly reduce edge-effects when the output tiles are later mosaicked together. When run in this batch mode, the output file (output) also need not be specified; the tool will instead create an output file with the same name as each input LiDAR file, but with the .tif extension. This can provide a very efficient means for processing extremely large LiDAR data sets.

Users may also exclude points from the interpolation if they fall below or above the minimum (minz) or maximum (maxz) thresholds respectively. This can be a useful means of excluding anomalously high or low points. Note that points that are classified as low points (LAS class 7) or high noise (LAS class 18) are automatically excluded from the interpolation operation.

Triangulation will generally completely fill the convex hull containing the input point data. This can sometimes result in very long and narrow triangles at the edges of the data or connecting vertices on either side of void areas. In LiDAR data, these void areas are often associated with larger waterbodies, and triangulation can result in very unnatural interpolated patterns within these areas. To avoid this problem, the user may specify a the maximum allowable triangle edge length (max_triangle_edge_length) and all grid cells within triangular facets with edges larger than this threshold are simply assigned the NoData values in the output DSM. These NoData areas can later be better dealt with using the fill_missing_data tool after interpolation.

See Also

lidar_tin_gridding, filter_lidar_classes, fill_missing_data, time_in_daylight

Function Signature

def lidar_digital_surface_model(self, input_lidar: Lidar, cell_size: float = 1.0, search_radius: float = 0.5, min_elev: float = float('-inf'), max_elev: float = float('inf'), max_triangle_edge_length: float = float('inf')) -> Raster: ...

lidar_elevation_slice

This tool can be used to either extract or classify the elevation values (z) of LiDAR points within a specified elevation range (slice). In addition to the names of the input and output LiDAR files (input and output), the user must specify the lower (minz) and upper (maxz) bounds of the elevation range. By default, the tool will only output points within the elevation slice, filtering out all points lying outside of this range. If the class parameter is used, the tool will operate by assigning a class value (inclassval) to the classification bit of points within the slice and another class value (outclassval) to those points falling outside the range.

See Also

lidar_remove_outliers, lidar_classify_subset

Function Signature

def lidar_elevation_slice(self, input: Lidar, minz: float = float('-inf'), maxz: float = float('inf'), classify: bool = False, in_class_value: int = 2, out_class_value: int = 1) -> Lidar: ...

lidar_ground_point_filter

This tool can be used to perform a slope-based classification, or filtering (i.e. removal), of non-ground points within a LiDAR point-cloud. The user must specify the name of the input and output LiDAR files (input and output). Inter-point slopes are compared between pair of points contained within local neighbourhoods of size radius. Neighbourhoods with fewer than the user-specified minimum number of points (min_neighbours) are extended until the minimum point number is equaled or exceeded. Points that are above neighbouring points by the minimum (height_threshold) and have an inter-point slope greater than the user-specifed threshold (slope_threshold) are considered non-ground points and are either optionally (classify) excluded from the output point-cloud or assigned the unclassified (value 1) class value.

Slope-based ground-point classification methods suffer from the challenge of uses a constant slope threshold under varying terrain slopes. Some researchers have developed schemes for varying the slope threshold based on underlying terrain slopes. lidar_ground_point_filter instead allow the user to optionally (slope_norm) normalize the underlying terrain (i.e. flatten the terrain) using a white top-hat transform. A constant slope threshold may then be used without contributing to poorer performance under steep topography. Note, that this option, while useful in rugged terrain, is computationally intensive. If the point-cloud is of a relatively flat terrain, this option may be excluded.

While this tool is appropriately applied to LiDAR point-clouds, the remove_off_terrain_objects tool can be used to remove off-terrain objects from rasterized LiDAR digital elevation models (DEMs).

Reference

Vosselman, G. (2000). Slope based filtering of laser altimetry data. International Archives of Photogrammetry and Remote Sensing, 33(B3/2; PART 3), 935-942.

See Also

remove_off_terrain_objects

Function Signature

def lidar_ground_point_filter(self, input_lidar: Lidar, search_radius: float = 2.0, min_neighbours: int = 0, slope_threshold: float = 45.0, height_threshold: float = 1.0, classify: bool = False, slope_norm: bool = True, height_above_ground: bool = False) -> Lidar: ...

lidar_hex_bin

The practice of binning point data to form a type of 2D histogram, density plot, or what is sometimes called a heatmap, is quite useful as an alternative for the cartographic display of of very dense points sets. This is particularly the case when the points experience significant overlap at the displayed scale. The lidar_point_density tool can be used to perform binning based on a regular grid (raster output). This tool, by comparison, bases the binning on a hexagonal grid.

The tool is similar to the CreateHexagonalVectorGrid tool, however instead will create an output hexagonal grid in which each hexagonal cell possesses a COUNT attribute which specifies the number of points from an input points file (LAS file) that are contained within the hexagonal cell. The tool will also calculate the minimum and maximum elevations and intensity values and outputs these data to the attribute table.

In addition to the names of the input points file and the output Shapefile, the user must also specify the desired hexagon width (w), which is the distance between opposing sides of each hexagon. The size (s) each side of the hexagon can then be calculated as, s = w / [2 x cos(PI / 6)]. The area of each hexagon (A) is, A = 3s(w / 2). The user must also specify the orientation of the grid with options of horizontal (pointy side up) and vertical (flat side up).

See Also

vector_hex_binning, lidar_point_density, CreateHexagonalVectorGrid

Function Signature

def lidar_hex_bin(self, input_lidar: Lidar, width: float, orientation: str = "h") -> Vector: ...

lidar_hillshade

Function Signature

def lidar_hillshade(self, input: Lidar, search_radius: float = -1.0, azimuth: float = 315.0, altitude: float = 30.0) -> Lidar: ...

lidar_histogram

This tool can be used to plot a histogram of data derived from a LiDAR file. The user must specify the name of the input LAS file (input), the name of the output HTML file (output), the parameter (parameter) to be plotted, and the amount (in percent) to clip the upper and lower tails of the f requency distribution (clip). The LiDAR parameters that can be plotted using lidar_histogram include the point elevations, intensity values, scan angles, and class values.

Use the lidar_point_stats tool instead to examine the spatial distribution of LiDAR points.

See Also

lidar_point_stats

Function Signature

def lidar_histogram(self, input_lidar: Lidar, output_html_file: str, parameter: str = "elevation", clip_percent: float = 1.0) -> None: ...

lidar_idw_interpolation

This tool interpolates LiDAR files using inverse-distance weighting (IDW) scheme. The user must specify the value of the IDW weight parameter (weight). The output grid can be based on any of the stored LiDAR point parameters (parameter), including elevation (in which case the output grid is a digital elevation model, DEM), intensity, class, return number, number of returns, scan angle, RGB (colour) values, and user data values. Similarly, the user may specify which point return values (returns) to include in the interpolation, including all points, last returns (including single return points), and first returns (including single return points).

The user must specify the grid resolution of the output raster (resolution), and optionally, the name of the input LiDAR file (input) and output raster (output). Note that if an input LiDAR file (input) is not specified by the user, the tool will search for all valid LiDAR (*.las, *.laz, *.zlidar) files contained within the current working directory. This feature can be very useful when you need to interpolate a DEM for a large number of LiDAR files. Not only does this batch processing mode enable the tool to run in a more optimized parallel manner, but it will also allow the tool to include a small buffer of points extending into adjacent tiles when interpolating an individual file. This can significantly reduce edge-effects when the output tiles are later mosaicked together. When run in this batch mode, the output file (output) also need not be specified; the tool will instead create an output file with the same name as each input LiDAR file, but with the .tif extension. This can provide a very efficient means for processing extremely large LiDAR data sets.

Users may excluded points from the interpolation based on point classification values, which follow the LAS classification scheme. Excluded classes are specified using the exclude_cls parameter. For example, to exclude all vegetation and building classified points from the interpolation, use --exclude_cls='3,4,5,6'. Users may also exclude points from the interpolation if they fall below or above the minimum (minz) or maximum (maxz) thresholds respectively. This can be a useful means of excluding anomalously high or low points. Note that points that are classified as low points (LAS class 7) or high noise (LAS class 18) are automatically excluded from the interpolation operation.

The tool will search for the nearest input LiDAR point to each grid cell centre, up to a maximum search distance (radius). If a grid cell does not have a LiDAR point within this search distance, it will be assigned the NoData value in the output raster. In LiDAR data, these void areas are often associated with larger waterbodies. These NoData areas can later be better dealt with using the fill_missing_data tool after interpolation.

See Also

lidar_tin_gridding, lidar_nearest_neighbour_gridding, LidarSibsonInterpolation

Function Signature

def lidar_idw_interpolation(self, input_lidar: Lidar, interpolation_parameter: str = "elevation", returns_included: str = "all", cell_size: float = 1.0, idw_weight: float = 1.0, search_radius: float = 2.5, excluded_classes: List[int] = None, min_elev: float = float('-inf'), max_elev: float = float('inf')) -> Raster: ...

lidar_info

This tool can be used to print basic information about the data contained within a LAS file, used to store LiDAR data. The reported information will include including data on the header, point return frequency, and classification data and information about the variable length records (VLRs) and geokeys.

Function Signature

def lidar_info(self, input_lidar: Lidar, output_html_file: str, show_point_density: bool = True, show_vlrs: bool = True, show_geokeys: bool = True) -> None: ...

lidar_join

This tool can be used to merge multiple LiDAR LAS files into a single output LAS file. Due to their large size, LiDAR data sets are often tiled into smaller, non-overlapping tiles. Sometimes it is more convenient to combine multiple tiles together for data processing and lidar_join can be used for this purpose.

See Also

lidar_tile

Function Signature

def lidar_join(self, inputs: List[Lidar]) -> Lidar: ...

lidar_kappa

This tool performs a kappa index of agreement (KIA) analysis on the classification values of two LiDAR (LAS) files. The output report HTML file should be displayed automatically but can also be displayed afterwards in any web browser. As a measure of overall classification accuracy, the KIA is more robust than the percent agreement calculation because it takes into account the agreement occurring by random chance. In addition to the KIA, the tool will output the producer's and user's accuracy, the overall accuracy, and the error matrix. The KIA is often used as a means of assessing the accuracy of an image classification analysis; however the LidarKappaIndex tool performs the analysis on a point-to-point basis, comparing the class values of the points in one input LAS file with the corresponding nearest points in the second input LAS file.

The user must also specify the name and resolution of an output raster file, which is used to show the spatial distribution of class accuracy. Each grid cell contains the overall accuracy, i.e. the points correctly classified divided by the total number of points contained within the cell, expressed as a percentage.

Function Signature

def lidar_kappa(self, input_lidar1: Lidar, input_lidar2: Lidar, output_html_file: str, cell_size: float = 1.0, output_class_accuracy: bool = False) -> Raster: ...

lidar_nearest_neighbour_gridding

This tool grids LiDAR files using nearest-neighbour (NN) scheme, that is, each grid cell in the output image will be assigned the parameter value of the point nearest the grid cell centre. This method should not be confused for the similarly named natural-neighbour interpolation (a.k.a Sibson's method). Nearest neighbour gridding is generally regarded as a poor way of interpolating surfaces from low-density point sets and results in the creation of a Voronoi diagram. However, this method has several advantages when applied to LiDAR data. NN gridding is one of the fastest methods for generating raster surfaces from large LiDAR data sets. NN gridding is one of the few interpolation methods, along with triangulation, that will preserve vertical breaks-in-slope, such as occur at the edges of building. This characteristic can be important when using some post-processing methods, such as the remove_off_terrain_objects tool. Furthermore, because most LiDAR data sets have remarkably high point densities compared with other types of geographic data, this approach does often produce a satisfactory result; this is particularly true when the point density is high enough that there are multiple points in the majority of grid cells.

The output grid can be based on any of the stored LiDAR point parameters (parameter), including elevation (in which case the output grid is a digital elevation model, DEM), intensity, class, return number, number of returns, scan angle, RGB (colour) values, and user data values. Similarly, the user may specify which point return values (returns) to include in the interpolation, including all points, last returns (including single return points), and first returns (including single return points).

The user must specify the grid resolution of the output raster (resolution), and optionally, the name of the input LiDAR file (input) and output raster (output). Note that if an input LiDAR file (input) is not specified by the user, the tool will search for all valid LiDAR (*.las, *.laz, *.zlidar) files contained within the current working directory. This feature can be very useful when you need to interpolate a DEM for a large number of LiDAR files. Not only does this batch processing mode enable the tool to run in a more optimized parallel manner, but it will also allow the tool to include a small buffer of points extending into adjacent tiles when interpolating an individual file. This can significantly reduce edge-effects when the output tiles are later mosaicked together. When run in this batch mode, the output file (output) also need not be specified; the tool will instead create an output file with the same name as each input LiDAR file, but with the .tif extension. This can provide a very efficient means for processing extremely large LiDAR data sets.

Users may excluded points from the interpolation based on point classification values, which follow the LAS classification scheme. Excluded classes are specified using the exclude_cls parameter. For example, to exclude all vegetation and building classified points from the interpolation, use --exclude_cls='3,4,5,6'. Users may also exclude points from the interpolation if they fall below or above the minimum (minz) or maximum (maxz) thresholds respectively. This can be a useful means of excluding anomalously high or low points. Note that points that are classified as low points (LAS class 7) or high noise (LAS class 18) are automatically excluded from the interpolation operation.

The tool will search for the nearest input LiDAR point to each grid cell centre, up to a maximum search distance (radius). If a grid cell does not have a LiDAR point within this search distance, it will be assigned the NoData value in the output raster. In LiDAR data, these void areas are often associated with larger waterbodies. These NoData areas can later be better dealt with using the fill_missing_data tool after interpolation.

See Also

lidar_tin_gridding, lidar_idw_interpolation, lidar_tin_gridding, remove_off_terrain_objects, fill_missing_data

Function Signature

def lidar_nearest_neighbour_gridding(self, input_lidar: Lidar, interpolation_parameter: str = "elevation", returns_included: str = "all", cell_size: float = 1.0, search_radius: float = 2.5, excluded_classes: List[int] = None, min_elev: float = float('-inf'), max_elev: float = float('inf')) -> Raster: ...

lidar_point_density

Function Signature

def lidar_point_density(self, input_lidar: Lidar, returns_included: str = "all", cell_size: float = 1.0, search_radius: float = 2.5, excluded_classes: List[int] = None, min_elev: float = float('-inf'), max_elev: float = float('inf')) -> Raster: ...

lidar_point_stats

This tool creates several rasters summarizing the distribution of LiDAR points in a LAS data file. The user must specify the name of an input LAS file (input) and the output raster grid resolution (resolution). Additionally, the user must specify one or more of the possible output rasters to create using the various available flags, which include:

FlagMeaning
num_pointsNumber of points (returns) in each grid cell
num_pulsesNumber of pulses in each grid cell
avg_points_per_pulseAverage number of points per pulse in each grid cells
z_rangeElevation range within each grid cell
intensity_rangeIntensity range within each grid cell
predom_classPredominant class value within each grid cell

If no output raster flags are specified, all of the output rasters will be created. All output rasters will have the same base name as the input LAS file but will have a suffix that reflects the statistic type (e.g. _num_pnts, _num_pulses, _avg_points_per_pulse, etc.). Output files will be in the GeoTIFF (*.tif) file format.

When the input/output parameters are not specified, the tool works on all LAS files contained within the working directory.

Notes:

  1. The num_pulses output is actually the number of pulses with at lease one return; specifically it is the sum of the early returns (first and only) in a grid cell. In areas of low reflectance, such as over water surfaces, the system may have emitted a significantly higher pulse rate but far fewer returns are observed.
  2. The memory requirement of this tool is high, particulalry if the grid resolution is fine and the spatial extent is large.

See Also

lidar_block_minimum, lidar_block_maximum

Function Signature

def lidar_point_stats(self, input_lidar: Lidar, cell_size: float = 1.0, num_points: bool = False, num_pulses: bool = False, avg_points_per_pulse: bool = False, z_range: bool = False, intensity_range: bool = False, predominant_class: bool = False) : ...

lidar_radial_basis_function_interpolation

Function Signature

def lidar_radial_basis_function_interpolation(self, input_lidar: Lidar, interpolation_parameter: str = "elevation", returns_included: str = "all", cell_size: float = 1.0, num_points: int = 15, excluded_classes: List[int] = None, min_elev: float = float('-inf'), max_elev: float = float('inf'), func_type: str = "thinplatespline", poly_order: str = "none", weight: float = 0.1) -> Raster: ...

lidar_ransac_planes

This tool uses the random sample consensus (RANSAC) method to identify points within a LiDAR point cloud that belong to planar surfaces. RANSAC is a common method used in the field of computer vision to identify a subset of inlier points in a noisy data set containing abundant outlier points. Because LiDAR point clouds often contain vegetation points that do not form planar surfaces, this tool can be used to largely strip vegetation points from the point cloud, leaving behind the ground returns, buildings, and other points belonging to planar surfaces. If the classify flag is used, non-planar points will not be removed but rather will be assigned a different class (1) than the planar points (0).

The algorithm selects a random sample, of a specified size (num_samples) of the points from within the neighbourhood (radius) surrounding each LiDAR point. The sample is then used to parameterize a planar best-fit model. The distance between each neighbouring point and the plane is then evaluated; inliers are those neighbouring points within a user-specified distance threshold (threshold). Models with at least a minimum number of inlier points (model_size) are then accepted. This process of selecting models is iterated a number of user-specified times (num_iter).

One of the challenges with identifying planar surfaces in LiDAR point clouds is that these data are usually collected along scan lines. Therefore, each scan line can potentially yield a vertical planar surface, which is one reason that some vegetation points remain after applying the RANSAC plane-fitting method. To cope with this problem, the tool allows the user to specify a maximum planar slope (max_slope) parameter. Planes that have slopes greater than this threshold are rejected by the algorithm. This has the side-effect of removing building walls however.

References

Fischler MA and Bolles RC. 1981. Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography. Commun. ACM, 24(6):381–395.

See Also

lidar_segmentation, lidar_ground_point_filter

Function Signature

def lidar_ransac_planes(self, in_lidar: Lidar, search_radius: float = 2.0, num_iterations: int = 50, num_samples: int = 10, inlier_threshold: float = 0.15, acceptable_model_size: int = 30, max_planar_slope: float = 75.0, classify: bool = False, only_last_returns: bool = False) -> Lidar: ...

lidar_remove_outliers

This tool will filter out points from a LiDAR point cloud if the absolute elevation difference between a point and the averge elevation of its neighbourhood, calculated without the point, exceeds a threshold (elev_diff).

Function Signature

def lidar_remove_outliers(self, input: Lidar, search_radius: float = 2.0, elev_diff: float = 50.0, use_median: bool = False, classify: bool = False) -> Lidar: ...

lidar_rooftop_analysis

This tool can be used to identify roof segments in a LiDAR point cloud.

See Also

classify_buildings_in_lidar, clip_lidar_to_polygon

Function Signature

def lidar_rooftop_analysis(self, lidar_inputs: List[Lidar], building_footprints: Vector, search_radius: float = 2.0, num_iterations: int = 50, num_samples: int = 10, inlier_threshold: float = 0.15, acceptable_model_size: int = 30, max_planar_slope: float = 75.0, norm_diff_threshold: float = 2.0, azimuth: float = 180.0, altitude: float = 30.0) -> Vector: ...

lidar_segmentation

This tool can be used to segment a LiDAR point cloud based on differences in the orientation of fitted planar surfaces and point proximity. The algorithm begins by attempting to fit planar surfaces to all of the points within a user-specified radius (radius) of each point in the LiDAR data set. The planar equation is stored for each point for which a suitable planar model can be fit. A region-growing algorithm is then used to assign nearby points with similar planar models. Similarity is based on a maximum allowable angular difference (in degrees) between the two neighbouring points' plane normal vectors (norm_diff). The norm_diff parameter can therefore be thought of as a way of specifying the magnitude of edges mapped by the region-growing algorithm. By setting this value appropriately, it is possible to segment each facet of a building's roof. Segment edges for planar points may also be determined by a maximum allowable height difference (maxzdiff) between neighbouring points on the same plane. Points for which no suitable planar model can be fit are assigned to 'volume' (non-planar) segments (e.g. vegetation points) using a region-growing method that connects neighbouring points based solely on proximity (i.e. all volume points within radius distance are considered to belong to the same segment).

The resulting point cloud will have both planar segments (largely ground surfaces and building roofs and walls) and volume segments (largely vegetation). Each segment is assigned a random red-green-blue (RGB) colour in the output LAS file. The largest segment in any airborne LiDAR dataset will usually belong to the ground surface. This largest segment will always be assigned a dark-green RGB of (25, 120, 0) by the tool.

This tool uses the random sample consensus (RANSAC) method to identify points within a LiDAR point cloud that belong to planar surfaces. RANSAC is a common method used in the field of computer vision to identify a subset of inlier points in a noisy data set containing abundant outlier points. Because LiDAR point clouds often contain vegetation points that do not form planar surfaces, this tool can be used to largely strip vegetation points from the point cloud, leaving behind the ground returns, buildings, and other points belonging to planar surfaces. If the classify flag is used, non-planar points will not be removed but rather will be assigned a different class (1) than the planar points (0).

The algorithm selects a random sample, of a specified size (num_samples) of the points from within the neighbourhood (radius) surrounding each LiDAR point. The sample is then used to parameterize a planar best-fit model. The distance between each neighbouring point and the plane is then evaluated; inliers are those neighbouring points within a user-specified distance threshold (threshold). Models with at least a minimum number of inlier points (model_size) are then accepted. This process of selecting models is iterated a number of user-specified times (num_iter).

One of the challenges with identifying planar surfaces in LiDAR point clouds is that these data are usually collected along scan lines. Therefore, each scan line can potentially yield a vertical planar surface, which is one reason that some vegetation points may be assigned to planes during the RANSAC plane-fitting method. To cope with this problem, the tool allows the user to specify a maximum planar slope (max_slope) parameter. Planes that have slopes greater than this threshold are rejected by the algorithm. This has the side-effect of removing building walls however.

References

Fischler MA and Bolles RC. 1981. Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography. Commun. ACM, 24(6):381–395.

See Also

lidar_ransac_planes, lidar_ground_point_filter

Function Signature

def lidar_segmentation(self, in_lidar: Lidar, search_radius: float = 2.0, num_iterations: int = 50, num_samples: int = 10, inlier_threshold: float = 0.15, acceptable_model_size: int = 30, max_planar_slope: float = 75.0, norm_diff_threshold: float = 2.0, max_z_diff: float = 1.0, classes: bool = False, ground: bool = False) -> Lidar: ...

lidar_segmentation_based_filter

Function Signature

def lidar_segmentation_based_filter(self, in_lidar: Lidar, search_radius: float = 5.0, norm_diff_threshold: float = 2.0, max_z_diff: float = 1.0, classify_points: bool = False) -> Lidar: ...

lidar_shift

This tool can be used to shift the x,y,z coordinates of points within a LiDAR file. The user must specify the name of the input file (input) and the output file (output). Additionally, the user must specify the x,y,z shift values (x_shift, y_shift, z_shift). At least one non-zero shift value is needed to run the tool. Notice that shifting the x,y,z coordinates of LiDAR points is also possible using the ModifyLidar tool, which can also be used for more sophisticated point property manipulation (e.g. rotations).

See Also

ModifyLidar, lidar_elevation_slice, height_above_ground

Function Signature

def lidar_shift(self, input: Lidar, x_shift: float = 0.0, y_shift: float = 0.0, z_shift: float = 0.0) -> Lidar: ...

lidar_thin

Thins a LiDAR point cloud, reducing point density.

Function Signature

def lidar_thin(self, input: Lidar, resolution: float = 1.0, selection_method: str = "first", save_filtered: bool = False) -> Tuple[Lidar, Union[Lidar, None]]: ...

lidar_thin_high_density

Thins points from high density areas within a LiDAR point cloud.

Function Signature

def lidar_thin_high_density(self, input: Lidar, density: float, resolution: float = 1.0, save_filtered: bool = False) -> Tuple[Lidar, Union[Lidar, None]]: ...

lidar_tile

single LAS file. The user must specify the parameter of the tile grid, including its origin (origin_x and origin_y) and the tile width and height (width and height). Tiles containing fewer points than specified in the min_points parameter will not be output. This can be useful when tiling terrestrial LiDAR datasets because the low point density at the edges of the point cloud (i.e. most distant from the scan station) can result in poorly populated tiles containing relatively few points.

See Also

lidar_join, lidar_tile_footprint

Function Signature

def lidar_tile(self, input_lidar: Lidar, tile_width: float = 1000.0, tile_height: float = 1000.0, origin_x: float = 0.0, origin_y: float = 0.0, min_points_in_tile: int = 2, output_laz_format: bool = True) -> None: ...

lidar_tile_footprint

This tool can be used to create a vector polygon of the bounding box or convex hull of a LiDAR point cloud (i.e. LAS file). If the user specified an input file (input) and output file (output), the tool will calculate the footprint, containing all of the data points, and output this feature to a vector polygon file. If the input and output parameters are left unspecified, the tool will calculate the footprint of every LAS file contained within the working directory and output these features to a single vector polygon file. If this is the desired mode of operation, it is important to specify the working directory (wd) containing the group of LAS files; do not specify the optional input and output parameters in this case. Each polygon in the output vector will contain a LAS_NM field, specifying the source LAS file name, a NUM_PNTS field, containing the number of points within the source file, and Z_MIN and Z_MAX fields, containing the minimum and maximum elevations. This output can therefore be useful to create an index map of a large tiled LiDAR dataset.

By default, this tool identifies the axis-aligned minimum rectangular hull, or bounding box, containing the points in each of the input tiles. If the user specifies the hull flag, the tool will identify the minimum convex hull instead of the bounding box. This option is considerably more computationally intensive and will be a far longer running operation if many tiles are specified as inputs.

A note on LAZ file inputs: While WhiteboxTools does not currently support the reading and writing of the compressed LiDAR format LAZ, it is able to read LAZ file headers. This tool, when run in in the bounding box mode (rather than the convex hull mode), is able to take LAZ input files.

lidar_tile, LayerFootprint, minimum_bounding_box, minimum_convex_hull

Function Signature

def lidar_tile_footprint(self, input_lidar: Lidar, output_hulls: bool = False) -> Vector: ...

lidar_tin_gridding

This tool creates a raster grid based on a Delaunay triangular irregular network (TIN) fitted to LiDAR points. The output grid can be based on any of the stored LiDAR point parameters (parameter), including elevation (in which case the output grid is a digital elevation model, DEM), intensity, class, return number, number of returns, scan angle, RGB (colour) values, and user data values. Similarly, the user may specify which point return values (returns) to include in the interpolation, including all points, last returns (including single return points), and first returns (including single return points).

The user must specify the grid resolution of the output raster (resolution), and optionally, the name of the input LiDAR file (input) and output raster (output). Note that if an input LiDAR file (input) is not specified by the user, the tool will search for all valid LiDAR (*.las, *.laz, *.zlidar) files contained within the current working directory. This feature can be very useful when you need to interpolate a DEM for a large number of LiDAR files. Not only does this batch processing mode enable the tool to run in a more optimized parallel manner, but it will also allow the tool to include a small buffer of points extending into adjacent tiles when interpolating an individual file. This can significantly reduce edge-effects when the output tiles are later mosaicked together. When run in this batch mode, the output file (output) also need not be specified; the tool will instead create an output file with the same name as each input LiDAR file, but with the .tif extension. This can provide a very efficient means for processing extremely large LiDAR data sets.

Users may excluded points from the interpolation based on point classification values, which follow the LAS classification scheme. Excluded classes are specified using the exclude_cls parameter. For example, to exclude all vegetation and building classified points from the interpolation, use --exclude_cls='3,4,5,6'. Users may also exclude points from the interpolation if they fall below or above the minimum (minz) or maximum (maxz) thresholds respectively. This can be a useful means of excluding anomalously high or low points. Note that points that are classified as low points (LAS class 7) or high noise (LAS class 18) are automatically excluded from the interpolation operation.

Triangulation will generally completely fill the convex hull containing the input point data. This can sometimes result in very long and narrow triangles at the edges of the data or connecting vertices on either side of void areas. In LiDAR data, these void areas are often associated with larger waterbodies, and triangulation can result in very unnatural interpolated patterns within these areas. To avoid this problem, the user may specify a the maximum allowable triangle edge length (max_triangle_edge_length) and all grid cells within triangular facets with edges larger than this threshold are simply assigned the NoData values in the output DSM. These NoData areas can later be better dealt with using the fill_missing_data tool after interpolation.

See Also

lidar_idw_interpolation, lidar_nearest_neighbour_gridding, lidar_tin_gridding, filter_lidar_classes, fill_missing_data

Function Signature

def lidar_tin_gridding(self, input_lidar: Lidar, interpolation_parameter: str = "elevation", returns_included: str = "all", cell_size: float = 1.0, excluded_classes: List[int] = None, min_elev: float = float('-inf'), max_elev: float = float('inf'), max_triangle_edge_length: float = float('inf')) -> Raster: ...

lidar_tophat_transform

This tool performs a white top-hat transform on a LiDAR point cloud (input). A top-hat transform is a common digital image processing operation used for various tasks, such as feature extraction, background equalization, and image enhancement. When applied to a LiDAR point cloud, the white top-hat transform provides an estimate of height above ground, which is useful for modelling the vegetation canopy.

As an example, notice that the input point cloud on the top of the image below has a substantial amount of topographic variability. After applying the top-hat transform (bottom point cloud), all of this topographic variability has been removed and point elevations values effectively become height above ground.

The white top-hat transform is defined as the difference between a point's original elevation and its opening. The opening operation can be thought of as the local neighbourhood maximum of a previous local minimum surface. The user must specify the size of the neighbourhood using the radius parameter. Setting this parameter can require some experimentation. Generally, it is appropriate to use a radius of a few meters in non-urban landscapes. However, in urban areas, the radius may need to be set much larger, reflective of the size of the largest building.

If the input point cloud already has ground points classified, it may be better to use the height_above_ground, which simply measures the difference in height between each point and its nearest ground classified point within the search radius.

See Also

height_above_ground, tophat_transform, closing, opening

Function Signature

def lidar_tophat_transform(self, input: Lidar, search_radius: float) -> Lidar: ...

line_detection_filter

This tool can be used to perform one of four 3x3 line-detection filters on a raster image. These filters can be used to find one-cell-thick vertical, horizontal, or angled (135-degrees or 45-degrees) lines in an image. Notice that line-finding is a similar application to edge-detection. Common edge-detection filters include the Sobel and Prewitt filters. The kernel weights for each of the four line-detection filters are as follows:

'v' (Vertical)

...
-12-1
-12-1
-12-1

'h' (Horizontal)

...
-1-1-1
222
-1-1-1

'45' (Northeast-Southwest)

...
-1-12
-12-1
2-1-1

'135' (Northwest-Southeast)

...
2-1-1
-12-1
-1-12

The user must specify the variant, including 'v', 'h', '45', and '135', for vertical, horizontal, northeast-southwest, and northwest-southeast directions respectively. The user may also optionally clip the output image distribution tails by a specified amount (e.g. 1%).

See Also

prewitt_filter, sobel_filter

Function Signature

def line_detection_filter(self, raster: Raster, variant: str = "v", abs_values: bool = False, clip_tails: float = 0.0) -> Raster: ...

line_intersections

This tool identifies points where the features of two vector line/polygon layers intersect. The user must specify the names of two input vector line files and the output file. The output file will be a vector of POINT VectorGeometryType. If the input vectors intersect at a line segment, the beginning and end vertices of the segment will be present in the output file. A warning is issued if intersection line segments are identified during analysis. If no intersections are found between the input line files, the output file will not be saved and a warning will be issued.

Each intersection point will contain PARENT1 and PARENT2 attribute fields, identifying the instersecting features in the first and second input line files respectively. Additionally, the output attribute table will contain all of the attributes (excluding FIDs) of the two parent line features.

Function Signature

def line_intersections(self, input1: Vector, input2: Vector) -> Vector: ...

line_thinning

This image processing tool reduces all polygons in a Boolean raster image to their single-cell wide skeletons. This operation is sometimes called line thinning or skeletonization. In fact, the input image need not be truly Boolean (i.e. contain only 1's and 0's). All non-zero, positive values are considered to be foreground pixels while all zero valued cells are considered background pixels. The remove_spurs tool is useful for cleaning up an image before performing a line thinning operation.

Note: Unlike other filter-based operations in WhiteboxTools, this algorithm can't easily be parallelized because the output raster must be read and written to during the same loop.

See Also

remove_spurs, thicken_raster_line

Function Signature

def line_thinning(self, raster: Raster) -> Raster: ...

linearity_index

This tool calculates the linearity index of polygon features based on a regression analysis. The index is simply the coefficient of determination (r-squared) calculated from a regression analysis of the x and y coordinates of the exterior hull nodes of a vector polygon. Linearity index is a measure of how well a polygon can be described by a straight line. It is a related index to the elongation_ratio, but is more efficient to calculate as it does not require finding the minimum bounding box. The Pearson correlation coefficient between linearity index and the elongation ratio for a large data set of lake polygons in northern Canada was found to be 0.656, suggesting a moderate level of association between the two measures of polygon linearity. Note that this index is not useful for identifying narrow yet sinuous polygons, such as meandering rivers.

The only required input is the name of the file. The linearity values calculated for each vector polygon feature will be placed in the accompanying attribute table as a new field (LINEARITY).

See Also

elongation_ratio, patch_orientation

Function Signature

def linearity_index(self, input: Vector) -> Vector: ...

lines_to_polygons

This tool converts vector polylines into polygons. Note that this tool will close polygons that are open and will ensure that the first part of an input line is interpreted as the polygon hull and subsequent parts are considered holes. The tool does not examine input lines for line crossings (self intersections), which are topological errors.

See Also

polygons_to_lines

Function Signature

def lines_to_polygons(self, input: Vector) -> Vector: ...

list_unique_values

This tool can be used to list each of the unique values contained within a categorical field of an input vector file's attribute table. The tool outputs an HTML formatted report (output) containing a table of the unique values and their frequency of occurrence within the data. The user must specify the name of an input shapefile (input) and the name of one of the fields (field) contained in the associated attribute table. The specified field should not contained floating-point numerical data, since the number of categories will likely equal the number of records, which may be quite large. The tool effectively provides tabular output that is similar to the graphical output provided by the attribute_histogram tool, which, however, can be applied to continuous data.

See Also

attribute_histogram

Function Signature

def list_unique_values(self, input: Vector, field_name: str) -> Tuple[str, int]: ...

long_profile

This tool can be used to create a longitudinal profile plot. A longitudinal stream profile is a plot of elevation against downstream distance. Most long profiles use distance from channel head as the distance measure. This tool, however, uses the distance to the stream network outlet cell, or mouth, as the distance measure. The reason for this difference is that while for any one location within a stream network there is only ever one downstream outlet, there are usually many upstream channel heads. Thus plotted using the traditional downstream-distance method, the same point within a network will plot in many different long profile locations, whereas it will always plot on one unique location in the distance-to-mouth method. One consequence of this difference is that the long profile will be oriented from right-to-left rather than left-to-right, as would traditionally be the case.

The tool outputs an interactive SVG line graph embedded in an HTML document (output_html_file). The user must input a D8 pointer (flow direction) raster (d8_pointer), a streams raster image (streams_raster), and a digital elevation model (dem). Stream cells are designated in the streams image as all positive, nonzero values. Thus all non-stream or background grid cells are commonly assigned either zeros or NoData values. The pointer image is used to traverse the stream network and should only be created using the D8 algorithm (d8_pointer). The streams image should be derived using a flow accumulation based stream network extraction algorithm, also based on the D8 flow algorithm.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, set esri_pointer=True.

See Also

long_profile_from_points, profile, d8_pointer

Function Signature

def long_profile(self, d8_pointer: Raster, streams_raster: Raster, dem: Raster, output_html_file: str, esri_pointer: bool = False) -> None: ...

long_profile_from_points

This tool can be used to create a longitudinal profile plot for a set of vector points (points). A longitudinal stream profile is a plot of elevation against downstream distance. Most long profiles use distance from channel head as the distance measure. This tool, however, uses the distance to the outlet cell, or mouth, as the distance measure.

The tool outputs an interactive SVG line graph embedded in an HTML document (output_html_file). The user input a D8 pointer (d8_pointer) image (flow direction), a vector points file (points), and a digital elevation model (dem). The pointer image is used to traverse the flow path issuing from each initiation point in the vector file; this pointer file should only be created using the D8 algorithm (d8_pointer).

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, the esri_pointer parameter must be specified.

See Also

long_profile, profile, d8_pointer

Function Signature

def long_profile_from_points(self, d8_pointer: Raster, points: Vector, dem: Raster, output_html_file: str, esri_pointer: bool = False) -> None: ...

longest_flowpath

This tool delineates the longest flowpaths for a group of subbasins or watersheds. Flowpaths are initiated along drainage divides and continue along the D8-defined flow direction until either the subbasin outlet or DEM edge is encountered. Each input subbasin/watershed will have an associated vector flowpath in the output image. longest_flowpath is similar to the r.lfp plugin tool for GRASS GIS. The length of the longest flowpath draining to an outlet is related to the time of concentration, which is a parameter used in certain hydrological models.

The user must input the filename of a digital elevation model (DEM), a basins raster, and the output vector. The DEM must be depressionless and should have been pre-processed using the breach_depressions_least_cost or fill_depressions tool. The basins raster must contain features that are delineated by categorical (integer valued) unique identifier values. All non-NoData, non-zero valued grid cells in the basins raster are interpreted as belonging to features. In practice, this tool is usual run using either a single watershed, a group of contiguous non-overlapping watersheds, or a series of nested subbasins. These are often derived using the watershed tool, based on a series of input outlets, or the subbasins tool, based on an input stream network. If subbasins are input to longest_flowpath, each traced flowpath will include only the non-overlapping portions within nested areas. Therefore, this can be a convenient method of delineating the longest flowpath to each bifurcation in a stream network.

The output vector file will contain fields in the attribute table that identify the associated basin unique identifier (BASIN), the elevation of the flowpath source point on the divide (UP_ELEV), the elevation of the outlet point (DN_ELEV), the length of the flowpath (LENGTH), and finally, the average slope (AVG_SLOPE) along the flowpath, measured as a percent grade.

See Also

max_upslope_flowpath_length, breach_depressions_least_cost, fill_depressions, watershed, subbasins

Function Signature

def longest_flowpath(self, dem: Raster, basins: Raster) -> Vector: ...

lowest_position

This tool identifies the stack position (index) of the minimum value within a raster stack on a cell-by-cell basis. For example, if five raster images (inputs) are input to the tool, the output raster (output) would show which of the five input rasters contained the lowest value for each grid cell. The index value in the output raster is the zero-order number of the raster stack, i.e. if the lowest value in the stack is contained in the first image, the output value would be 0; if the lowest stack value were the second image, the output value would be 1, and so on. If any of the cell values within the stack is NoData, the output raster will contain the NoData value for the corresponding grid cell. The index value is related to the order of the input images.

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

highest_position, pick_from_list

Function Signature

def lowest_position(self, input_rasters: List[Raster]) -> Raster: ...

majority_filter

This tool performs a range filter on an input image (input). A range filter assigns to each cell in the output grid. The range (maximum - minimum) of the values contained within a moving window centred on each grid cell.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

See Also

total_filter

Function Signature

def majority_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

map_off_terrain_objects

This tool can be used to map off-terrain objects in a digital surface model (DSM) based on cell-to-cell differences in elevations and local slopes. The algorithm works by using a region-growing operation to connect neighbouring grid cells outwards from seed cells. Two neighbouring cells are considered connected if the slope between the two cells is less than the user-specified maximum slope value (max_slope). Mapped segments that are less than the minimum feature size (min_size), in grid cells, are assigned a common background value. Note that this method of mapping off-terrain objects, and thereby separating ground cells from non-ground objects in DSMs, works best with fine-resolution DSMs that have been interpolated using a non-smoothing method, such as triangulation (TINing) or nearest-neighbour interpolation.

See Also

remove_off_terrain_objects

Function Signature

def map_off_terrain_objects(self, dem: Raster, max_slope: float = float('inf'), min_feature_size: int = 0) -> Raster: ...

max_absolute_overlay

This tool can be used to find the maximum absolute (non-negative) value in each cell of a grid from a set of input images (inputs). NoData values in any of the input images will result in a NoData pixel in the output image.

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

max_overlay, min_absolute_overlay, min_overlay

Function Signature

def max_absolute_overlay(self, input_rasters: List[Raster]) -> Raster: ...

max_anisotropy_dev

Calculates the maximum anisotropy (directionality) in elevation deviation over a range of spatial scales.

Function Signature

def max_anisotropy_dev(self, dem: Raster, min_scale: int = 1, max_scale: int = 100, step_size: int = 1) -> Tuple[Raster, Raster]: ...

max_anisotropy_dev_signature

/.//

Function Signature

def max_anisotropy_dev_signature(self, dem: Raster, points: Vector, output_html_file: str, min_scale: int = 1, max_scale: int = 100, step_size: int = 1) -> None: ...

max_branch_length

Maximum branch length (Bmax) is the longest branch length between a grid cell's flowpath and the flowpaths initiated at each of its neighbours. It can be conceptualized as the downslope distance that a volume of water that is split into two portions by a drainage divide would travel before reuniting.

If the two flowpaths of neighbouring grid cells do not intersect, Bmax is simply the flowpath length from the starting cell to its terminus at the edge of the grid or a cell with undefined flow direction (i.e. a pit cell either in a topographic depression or at the edge of a major body of water).

The pattern of Bmax derived from a DEM should be familiar to anyone who has interpreted upslope contributing area images. In fact, Bmax can be thought of as the complement of upslope contributing area. Whereas contributing area is greatest along valley bottoms and lowest at drainage divides, Bmax is greatest at divides and lowest along channels. The two topographic attributes are also distinguished by their units of measurements; Bmax is a length rather than an area. The presence of a major drainage divide between neighbouring grid cells is apparent in a Bmax image as a linear feature, often two grid cells wide, of relatively high values. This property makes Bmax a useful land surface parameter for mapping ridges and divides.

Bmax is useful in the study of landscape structure, particularly with respect to drainage patterns. The index gives the relative significance of a specific location along a divide, with respect to the dispersion of materials across the landscape, in much the same way that stream ordering can be used to assess stream size.

See Also

flow_length_diff

Reference

Lindsay JB, Seibert J. 2013. Measuring the significance of a divide to local drainage patterns. International Journal of Geographical Information Science, 27: 1453-1468. DOI: 10.1080/13658816.2012.705289

Function Signature

def max_branch_length(self, dem: Raster, log_transform: bool = False) -> Raster: ...

max_difference_from_mean

Calculates the maximum difference from mean elevation over a range of spatial scales.

Function Signature

def max_difference_from_mean(self, dem: Raster, min_scale: int = 1, max_scale: int = 100, step_size: int = 1) -> Tuple[Raster, Raster]: ...

max_downslope_elev_change

This tool calculates the maximum elevation drop between each grid cell and its neighbouring cells within a digital elevation model (DEM). The user must input a DEM (dem).

See Also

max_upslope_elev_change, min_downslope_elev_change, num_downslope_neighbours

Function Signature

def max_downslope_elev_change(self, raster: Raster) -> Raster: ...

max_elevation_dev_signature

max_elevation_dev_signature

Tool documentation not located.

Function Signature

def max_elevation_dev_signature(self, dem: Raster, points: Vector, output_html_file: str, min_scale: int = 1, max_scale: int = 100, step_size: int = 1) -> None: ...

max_elevation_deviation

This tool can be used to calculate the maximum deviation from mean elevation, DEVmax (Lindsay et al. 2015) for each grid cell in a digital elevation model (DEM) across a range specified spatial scales. DEV is an elevation residual index and is essentially equivalent to a local elevation z-score. This attribute measures the relative topographic position as a fraction of local relief, and so is normalized to the local surface roughness. The multi-scaled calculation of DEVmax utilizes an integral image approach (Crow, 1984) to ensure highly efficient filtering that is invariant with filter size, which is the algorithm characteristic that allows for this densely sampled multi-scale analysis. In this way, max_elevation_deviation allows users to estimate the locally optimal scale with which to estimate DEV on a pixel-by-pixel basis. This multi-scaled version of local topographic position can reveal significant terrain characteristics and can aid with soil, vegetation, landform, and other mapping applications that depend on geomorphometric characterization.

The user must input a digital elevation model (DEM) (dem). The range of scales that are evaluated in calculating DEVmax are determined by the user-specified min_scale, max_scale, and step parameters. All filter radii between the minimum and maximum scales, increasing by step, will be evaluated. The scale parameters are in units of grid cells and specify kernel size "radii" (r), such that:

d = 2r + 1

That is, a radii of 1, 2, 3... yields a square filters of dimension (d) 3 x 3, 5 x 5, 7 x 7...

DEV is estimated at each tested filter size and every grid cell is assigned the maximum DEV value across the evaluated scales.

Two output rasters will be generated, including the magnitude (DEVmax) and a second raster the assigns each pixel the scale at which DEVmax is encountered (DEVscale). The DEVscale raster can be very useful for revealing multi-scale landscape structure.

Reference

Lindsay J, Cockburn J, Russell H. 2015. An integral image approach to performing multi-scale topographic position analysis. Geomorphology, 245: 51-61.

See Also

DevFromMeanElev, max_difference_from_mean, multiscale_elevation_percentile

Function Signature

def max_elevation_deviation(self, dem: Raster, min_scale: int = 1, max_scale: int = 100, step_size: int = 1) -> Tuple[Raster, Raster]: ...

max_overlay

This tool can be used to find the maximum value in each cell of a grid from a set of input images (inputs). NoData values in any of the input images will result in a NoData pixel in the output image (output). It is similar to the Max mathematical tool, except that it will accept more than two input images.

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

min_overlay, max_absolute_overlay, min_absolute_overlay, Max

Function Signature

def max_overlay(self, input_rasters: List[Raster]) -> Raster: ...

max_procs

Determines the number of processors used by functions that are parallelized. If set to -1 (wbe.max_procs=-1), the default, all available processors will be used. To throttle tools, set max_procs to a positive whole number less than the number of system processors.

max_upslope_elev_change

a digital elevation model (DEM). The user must input DEM (dem).

See Also

max_downslope_elev_change

Function Signature

def max_upslope_elev_change(self, raster: Raster) -> Raster: ...

max_upslope_flowpath_length

This tool calculates the maximum length of the flowpaths that run through each grid cell (in map horizontal units) in an input digital elevation model (dem). The tool works by first calculating the D8 flow pointer (d8_pointer) from the input DEM. The DEM must be depressionless and should have been pre-processed using the breach_depressions_least_cost or fill_depressions tool. The user must also specify the name of output raster (output).

See Also

d8_pointer, breach_depressions_least_cost, fill_depressions, average_upslope_flowpath_length, downslope_flowpath_length, downslope_distance_to_stream

Function Signature

def max_upslope_flowpath_length(self, dem: Raster) -> Raster: ...

max_upslope_value

This tool calculates the maximum length of the flowpaths that run through each grid cell (in map horizontal units) in an input digital elevation model (dem). The tool works by first calculating the D8 flow pointer (d8_pointer) from the input DEM. The DEM must be depressionless and should have been pre-processed using the breach_depressions_least_cost or fill_depressions tool. The user must also specify the name of output raster (output).

See Also

d8_pointer, breach_depressions_least_cost, fill_depressions, average_upslope_flowpath_length, downslope_flowpath_length, downslope_distance_to_stream

Function Signature

def max_upslope_value(self, dem: Raster, values_raster: Raster) -> Raster: ...

maximal_curvature

This tool calculates the maximal curvature from a digital elevation model (DEM). Maximal curvature is the curvature of a principal section with the highest value of curvature at a given point of the topographic surface (Florinsky, 2017). The values of this curvature are unbounded, and positive values correspond to ridge positions while negative values are indicative of closed depressions (Florinsky, 2016). Maximal curvature is measured in units of m-1.

The user must input a DEM (dem). The Z conversion factor (zfactor) is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z Conversion Factor. Curvature values are often very small and as such the user may opt to log-transform the output raster (log). Transforming the values applies the equation by Shary et al. (2002):

Θ' = sign(Θ) ln(1 + 10n|Θ|)

where Θ is the parameter value and n is dependent on the grid cell size.

For DEMs in projected coordinate systems, the tool uses the 3rd-order bivariate Taylor polynomial method described by Florinsky (2016). Based on a polynomial fit of the elevations within the 5x5 neighbourhood surrounding each cell, this method is considered more robust against outlier elevations (noise) than other methods. For DEMs in geographic coordinate systems (i.e. angular units), the tool uses the 3x3 polynomial fitting method for equal angle grids also described by Florinsky (2016).

References

Florinsky, I. (2016). Digital terrain analysis in soil science and geology. Academic Press.

Florinsky, I. V. (2017). An illustrated introduction to general geomorphometry. Progress in Physical Geography, 41(6), 723-752.

Shary P. A., Sharaya L. S. and Mitusov A. V. (2002) Fundamental quantitative methods of land surface analysis. Geoderma 107: 1–32.

minimal_curvature, tangential_curvature, profile_curvature, plan_curvature, mean_curvature, gaussian_curvature

Function Signature

def maximal_curvature(self, dem: Raster, log_transform: bool = False, z_factor: float = 1.0) -> Raster: ...

maximum_filter

This tool assigns each cell in the output grid. The maximum value in a moving window centred on each grid cell in the input raster (input). A maximum filter is the equivalent of the mathematical morphological dilation operator.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values, e.g. 3, 5, 7, 9... If the kernel filter size is the same in the x and y dimensions, the silent filter flag may be used instead (command-line interface only).

This tool takes advantage of the redundancy between overlapping, neighbouring filters to enhance computationally efficiency. Like most of WhiteboxTools' filters, it is also parallelized for further efficiency.

See Also

minimum_filter

Function Signature

def maximum_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

mdinf_flow_accum

This tool is used to generate a flow accumulation grid (i.e. contributing area) using the MD-infinity algorithm (Seibert and McGlynn, 2007). This algorithm is an examples of a multiple-flow-direction (MFD) method because the flow entering each grid cell is routed to one or two downslope neighbour, i.e. flow divergence is permitted. The user must specify the name of the input digital elevation model (dem). The DEM should have been hydrologically corrected to remove all spurious depressions and flat areas. DEM pre-processing is usually achieved using the breach_depressions_least_cost or fill_depressions tool.

In addition to the input flow-pointer grid name, the user must specify the output type (out_type). The output flow-accumulation can be 1) specific catchment area (SCA), which is the upslope contributing area divided by the contour length (taken as the grid resolution), 2) total catchment area in square-metres, or 3) the number of upslope grid cells. The user must also specify whether the output flow-accumulation grid should be log-tranformed, i.e. the output, if this option is selected, will be the natural-logarithm of the accumulated area. This is a transformation that is often performed to better visualize the contributing area distribution. Because contributing areas tend to be very high along valley bottoms and relatively low on hillslopes, when a flow-accumulation image is displayed, the distribution of values on hillslopes tends to be 'washed out' because the palette is stretched out to represent the highest values. Log-transformation (log) provides a means of compensating for this phenomenon. Importantly, however, log-transformed flow-accumulation grids must not be used to estimate other secondary terrain indices, such as the wetness index, or relative stream power index.

Grid cells possessing the NoData value in the input DEM raster are assigned the NoData value in the output flow-accumulation image. The output raster is of the float data type and continuous data scale.

Reference

Seibert, J. and McGlynn, B.L., 2007. A new triangular multiple flow direction algorithm for computing upslope areas from gridded digital elevation models. Water resources research, 43(4).

See Also

D8FlowAccumulation, FD8FlowAccumulation, quinn_flow_accumulation, qin_flow_accumulation, DInfFlowAccumulation, MDInfFlowAccumulation, rho8_pointer, breach_depressions_least_cost

Function Signature

def mdinf_flow_accum(self, dem: Raster, out_type: str = "sca", exponent: float = 1.1, convergence_threshold: float = float('inf'), log_transform: bool = False, clip: bool = False) -> Raster: ...

mean_curvature

This tool calculates the mean curvature, or the rate of change in slope along a flow line, from a digital elevation model (DEM). Curvature is the second derivative of the topographic surface defined by a DEM. Profile curvature characterizes the degree of downslope acceleration or deceleration within the landscape (Gallant and Wilson, 2000). The user must input a DEM (dem). WhiteboxTools reports curvature in radians multiplied by 100 for easier interpretation because curvature values are typically very small. The Z conversion factor (zfactor) is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z Conversion Factor. If the DEM is in the geographic coordinate system (latitude and longitude), the following equation is used:

zfactor = 1.0 / (111320.0 x cos(mid_lat))

where mid_lat is the latitude of the centre of the raster, in radians.

The algorithm uses the same formula for the calculation of plan curvature as Gallant and Wilson (2000). Profile curvature is negative for slope increasing downhill (convex flow profile, typical of upper slopes) and positive for slope decreasing downhill (concave, typical of lower slopes).

Reference

Gallant, J. C., and J. P. Wilson, 2000, Primary topographic attributes, in Terrain Analysis: Principles and Applications, edited by J. P. Wilson and J. C. Gallant pp. 51-86, John Wiley, Hoboken, N.J.

See Also

profile_curvature, tangential_curvature, total_curvature, slope, aspect

Function Signature

def mean_curvature(self, dem: Raster, log_transform: bool = False, z_factor: float = 1.0) -> Raster: ...

mean_filter

This tool performs a mean filter operation on a raster image. A mean filter, a type of low-pass filter, can be used to emphasize the longer-range variability in an image, effectively acting to smooth the image. This can be useful for reducing the noise in an image. This tool utilizes an integral image approach (Crow, 1984) to ensure highly efficient filtering that is invariant to filter size. The algorithm operates by calculating the average value in a moving window centred on each grid cell. Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values, e.g. 3, 5, 7, 9... If the kernel filter size is the same in the x and y dimensions, the silent filter flag may be used instead (command-line interface only).

Although commonly applied in digital image processing, mean filters are generally considered to be quite harsh, with respect to their impact on the image, compared to other smoothing filters such as the edge-preserving smoothing filters including the bilateral_filter, median_filter, olympic_filter, edge_preserving_mean_filter and even gaussian_filter.

This tool works with both greyscale and red-green-blue (RGB) images. RGB images are decomposed into intensity-hue-saturation (IHS) and the filter is applied to the intensity channel. NoData values in the input image are ignored during filtering. NoData values are assigned to all sites beyond the raster.

Reference

Crow, F. C. (1984, January). Summed-area tables for texture mapping. In ACM SIGGRAPH computer graphics (Vol. 18, No. 3, pp. 207-212). ACM.

See Also

bilateral_filter, edge_preserving_mean_filter, gaussian_filter, median_filter, rgb_to_ihs

Function Signature

def mean_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

median_filter

This tool performs a median filter on a raster image. Median filters, a type of low-pass filter, can be used to emphasize the longer-range variability in an image, effectively acting to smooth the image. This can be useful for reducing the noise in an image. The algorithm operates by calculating the median value (middle value in a sorted list) in a moving window centred on each grid cell. Specifically, this tool uses the efficient running-median filtering algorithm of Huang et al. (1979). The median value is not influenced by anomolously high or low values in the distribution to the extent that the average is. As such, the median filter is far less sensitive to shot noise in an image than the mean filter.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filteryflags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

Reference

Huang, T., Yang, G.J.T.G.Y. and Tang, G., 1979. A fast two-dimensional median filtering algorithm. IEEE Transactions on Acoustics, Speech, and Signal Processing, 27(1), pp.13-18.

See Also

bilateral_filter, edge_preserving_mean_filter, gaussian_filter, mean_filter

Function Signature

def median_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11, sig_digits: int = 2) -> Raster: ...

medoid

This tool calculates the medoid for a series of vector features contained in a shapefile. The medoid of a two-dimensional feature is conceptually similar its centroid, or mean position, but the medoid is always a members of the input feature data set. Thus, the medoid is a measure of central tendency that is robust in the presence of outliers. If the input vector is of a POLYLINE or POLYGON VectorGeometryType, the nodes of each feature will be used to estimate the feature medoid. If the input vector is of a POINT base VectorGeometryType, the medoid will be calculated for the collection of points. While there are more than one competing method of calculating the medoid, this tool uses an algorithm that works as follows:

  1. The x-coordinate and y-coordinate of each point/node are placed into two arrays.
  2. The x- and y-coordinate arrays are then sorted and the median x-coordinate (Med X) and median y-coordinate (Med Y) are calculated.
  3. The point/node in the dataset that is nearest the point (Med X, Med Y) is identified as the medoid.

See Also

centroid_vector

Function Signature

def medoid(self, input: Vector) -> Vector: ...

merge_line_segments

Vector lines can sometimes contain two features that are connected by a shared end vertex. This tool identifies connected line features in an input vector file (input) and merges them in the output file (output). Two line features are merged if their ends are coincident, and are not coincident with any other feature (i.e. a bifurcation junction). End vertices are considered to be coincident if they are within the specified snap distance (snap).

See Also

split_with_lines

Function Signature

def merge_line_segments(self, input: Vector, snap_tolerance: float = 2.220446049250313e-16) -> Vector: ...

merge_table_with_csv

This tool can be used to merge a vector's attribute table with data contained within a comma separated values (CSV) text file. CSV files stores tabular data (numbers and text) in plain-text form such that each row is a record and each column a field. Fields are typically separated by commas although the tool will also support seimi-colon, tab, and space delimited files. The user must specify the name of the vector (and associated attribute file) as well as the primary key within the table. The primary key (pkey flag) is the field within the table that is being appended to that serves as the unique identifier. Additionally, the user must specify the name of a CSV text file with either a *.csv or *.txt extension. The file must possess a header row, i.e. the first row must contain information about the names of the various fields. The foreign key (fkey flag), that is the identifying field within the CSV file that corresponds with the data contained within the primary key in the table, must also be specified. Both the primary and foreign keys should either be strings (text) or integer values. Fields containing decimal values are not good candidates for keys. Lastly, the user may optionally specify the name of a field within the CSV file to import in the merge operation (import_field flag). If this flag is not specified, all of the fields within the CSV, with the exception of the foreign key, will be appended to the attribute table.

Merging works for one-to-one and many-to-one database relations. A one-to-one relations exists when each record in the attribute table corresponds to one record in the second table and each primary key is unique. Since each record in the attribute table is associated with a geospatial feature in the vector, an example of a one-to-one relation may be where the second file contains AREA and PERIMETER fields for each polygon feature in the vector. This is the most basic type of relation. A many-to-one relation would exist when each record in the first attribute table corresponds to one record in the second file and the primary key is NOT unique. Consider as an example a vector and attribute table associated with a world map of countries. Each country has one or more more polygon features in the shapefile, e.g. Canada has its mainland and many hundred large islands. You may want to append a table containing data about the population and area of each country. In this case, the COUNTRY columns in the attribute table and the second file serve as the primary and foreign keys respectively. While there may be many duplicate primary keys (all of those Canadian polygons) each will correspond to only one foreign key containing the population and area data. This is a many-to-one relation. The join_tables tool does not support one-to-many nor many-to-many relations.

See Also

join_tables, reinitialize_attribute_table, export_table_to_csv

Function Signature

def merge_table_with_csv(self, primary_vector: Vector, primary_key_field: str, foreign_csv_filename: str, foreign_key_field: str, import_field: str = "") -> None: ...

merge_vectors

Combines two or more input vectors of the same ShapeType creating a single, new output vector. Importantly, the attribute table of the output vector will contain the ubiquitous file-specific FID, the parent file name, the parent FID, and the list of attribute fields that are shared among each of the input files. For a field to be considered common between tables, it must have the same name and field_type (i.e. data type and precision).

Overlapping features will not be identified nor handled in the merging. If you have significant areas of overlap, it is advisable to use one of the vector overlay tools instead.

The difference between merge_vectors and the Append tool is that merging takes two or more files and creates one new file containing the features of all inputs, and Append places the features of a single vector into another existing (appended) vector.

This tool only operates on vector files. Use the mosaic tool to combine raster data.

See Also

Append, mosaic

Function Signature

def merge_vectors(self, input_vectors: List[Vector]) -> Vector: ...

min_absolute_overlay

This tool can be used to find the minimum absolute (non-negative) value in each cell of a grid from a set of input images (inputs). NoData values in any of the input images will result in a NoData pixel in the output image.

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

min_overlay, max_absolute_overlay, max_overlay

Function Signature

def min_absolute_overlay(self, input_rasters: List[Raster]) -> Raster: ...

min_downslope_elev_change

This tool calculates the minimum elevation drop between each grid cell and its neighbouring cells within a digital elevation model (DEM). The user must input a DEM (dem).

See Also

max_downslope_elev_change, num_downslope_neighbours

Function Signature

def min_downslope_elev_change(self, raster: Raster) -> Raster: ...

min_max_contrast_stretch

This tool performs a Gaussian stretch on a raster image. The observed histogram of the input image is fitted to a Gaussian histogram, i.e. normal distribution. A histogram matching technique is used to map the values from the input image onto the output Gaussian distribution. The user must the number of tones (num_tones) used.

This tool is related to the more general histogram_matching tool, which can be used to fit any frequency distribution to an input image, and other contrast enhancement tools such as histogram_equalization, min_max_contrast_stretch, percentage_contrast_stretch, sigmoidal_contrast_stretch, and standard_deviation_contrast_stretch.

See Also

PiecewiseContrastStretch, histogram_equalization, min_max_contrast_stretch, percentage_contrast_stretch, sigmoidal_contrast_stretch, standard_deviation_contrast_stretch, histogram_matching

Function Signature

def min_max_contrast_stretch(self, raster: Raster, min_val: float, max_val: float, num_tones: int = 256) -> Raster: ...

min_overlay

This tool can be used to find the minimum value in each cell of a grid from a set of input images (inputs). NoData values in any of the input images will result in a NoData pixel in the output image (output). It is similar to the Min mathematical tool, except that it will accept more than two input images.

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

max_overlay, max_absolute_overlay, min_absolute_overlay, Min

Function Signature

def min_overlay(self, input_rasters: List[Raster]) -> Raster: ...

minimal_curvature

This tool calculates the minimal curvature from a digital elevation model (DEM). Minimal curvature is the curvature of a principal section with the lowest value of curvature at a given point of the topographic surface (Florinsky, 2017). The values of this curvature are unbounded, and positive values correspond to hills while negative values are indicative of valley positions (Florinsky, 2016). Minimal curvature is measured in units of m-1.

The user must input a DEM (dem). The Z conversion factor (zfactor) is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z Conversion Factor. Curvature values are often very small and as such the user may opt to log-transform the output raster (log). Transforming the values applies the equation by Shary et al. (2002):

Θ' = sign(Θ) ln(1 + 10n|Θ|)

where Θ is the parameter value and n is dependent on the grid cell size.

For DEMs in projected coordinate systems, the tool uses the 3rd-order bivariate Taylor polynomial method described by Florinsky (2016). Based on a polynomial fit of the elevations within the 5x5 neighbourhood surrounding each cell, this method is considered more robust against outlier elevations (noise) than other methods. For DEMs in geographic coordinate systems (i.e. angular units), the tool uses the 3x3 polynomial fitting method for equal angle grids also described by Florinsky (2016).

References

Florinsky, I. (2016). Digital terrain analysis in soil science and geology. Academic Press.

Florinsky, I. V. (2017). An illustrated introduction to general geomorphometry. Progress in Physical Geography, 41(6), 723-752.

Shary P. A., Sharaya L. S. and Mitusov A. V. (2002) Fundamental quantitative methods of land surface analysis. Geoderma 107: 1–32.

maximal_curvature, tangential_curvature, profile_curvature, plan_curvature, mean_curvature, gaussian_curvature

Function Signature

def minimal_curvature(self, dem: Raster, log_transform: bool = False, z_factor: float = 1.0) -> Raster: ...

minimum_bounding_box

This tool delineates the minimum bounding box (MBB) for a group of vectors. The MBB is the smallest box to completely enclose a feature. The algorithm works by rotating the feature, calculating the axis-aligned bounding box for each rotation, and finding the box with the smallest area, length, width, or perimeter. The MBB is needed to compute several shape indices, such as the Elongation Ratio. The MinimumBoundingEnvelop tool can be used to calculate the axis-aligned bounding rectangle around each feature in a vector file.

See Also

minimum_bounding_circle, minimum_bounding_envelope, minimum_convex_hull

Function Signature

def minimum_bounding_box(self, input: Vector, min_criteria: str = "area", individual_feature_hulls: bool = True) -> Vector: ...

minimum_bounding_circle

This tool delineates the minimum bounding circle (MBC) for a group of vectors. The MBC is the smallest enclosing circle to completely enclose a feature.

See Also

minimum_bounding_box, minimum_bounding_envelope, minimum_convex_hull

Function Signature

def minimum_bounding_circle(self, input: Vector, individual_feature_hulls: bool = True) -> Vector: ...

minimum_bounding_envelope

This tool delineates the minimum bounding axis-aligned box for a group of vector features. The is the smallest rectangle to completely enclose a feature, in which the sides of the envelope are aligned with the x and y axis of the coordinate system. The minimum_bounding_box can be used instead to find the smallest possible non-axis aligned rectangular envelope.

See Also

minimum_bounding_box, minimum_bounding_circle, minimum_convex_hull

Function Signature

def minimum_bounding_envelope(self, input: Vector, individual_feature_hulls: bool = True) -> Vector: ...

minimum_convex_hull

This tool creates a vector convex polygon around vector features. The convex hull is a convex closure of a set of points or polygon vertices and can be may be conceptualized as the shape enclosed by a rubber band stretched around the point set. The convex hull has many applications and is most notably used in various shape indices. The Delaunay triangulation of a point set and its dual, the Voronoi diagram, are mathematically related to convex hulls.

See Also

minimum_bounding_box, minimum_bounding_circle, minimum_bounding_envelope

Function Signature

def minimum_convex_hull(self, input: Vector, individual_feature_hulls: bool = True) -> Vector: ...

minimum_filter

This tool assigns each cell in the output grid the minimum value in a moving window centred on each grid cell in the input raster (input). A maximum filter is the equivalent of the mathematical morphological erosion operator.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values, e.g. 3, 5, 7, 9... If the kernel filter size is the same in the x and y dimensions, the silent filter flag may be used instead (command-line interface only).

This tool takes advantage of the redundancy between overlapping, neighbouring filters to enhance computationally efficiency. Like most of WhiteboxTools' filters, it is also parallelized for further efficiency.

See Also

maximum_filter

Function Signature

def minimum_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

modified_k_means_clustering

This modified k-means algorithm is similar to that described by Mather and Koch (2011). The main difference between the traditional k-means and this technique is that the user does not need to specify the desired number of classes/clusters prior to running the tool. Instead, the algorithm initializes with a very liberal overestimate of the number of classes and then merges classes that have cluster centres that are separated by less than a user-defined threshold. The main difference between this algorithm and the ISODATA technique is that clusters can not be broken apart into two smaller clusters.

Reference

Mather, P. M., & Koch, M. (2011). Computer processing of remotely-sensed images: an introduction. John Wiley & Sons.

See Also

k_means_clustering

Function Signature

def modified_k_means_clustering(self, input_rasters: List[Raster], output_html_file: str = "", num_start_clusters: int = 1000, merge_distance: float = 1.0, max_iterations: int = 10, percent_changed_threshold: float = 2.0) -> Raster: ...

modify_nodata_value

This tool can be used to modify the value of pixels containing the NoData value for an input raster image. This operation differs from the set_nodata_value tool, which sets the NoData value for an image in the image header without actually modifying pixel values. Also, set_nodata_value does not overwrite the input file, while the modify_nodata_value tool does.

See Also

set_nodata_value, convert_nodata_to_zero

Function Signature

def modify_nodata_value(self, input: Raster, new_value: float = -32768.0) : ...

mosaic

This tool will create an image mosaic from one or more input image files using one of three resampling methods including, nearest neighbour, bilinear interpolation, and cubic convolution. The order of the input source image files is important. Grid cells in the output image will be assigned the corresponding value determined from the last image found in the list to possess an overlapping coordinate.

Note that when the inputs parameter is left unspecified, the tool will use all of the .tif, .tiff, .rdc, .flt, .sdat, and .dep files located in the working directory. This can be a useful way of mosaicing large number of tiles, particularly when the text string that would be required to specify all of the input tiles is longer than the allowable limit.

This is the preferred mosaicing tool to use when appending multiple images with little to no overlapping areas, e.g. tiled data. When images have significant overlap areas, users are advised to use the mosaic_with_feathering tool instead.

Resample is very similar in operation to the Mosaic tool. The Resample tool should be used when there is an existing image into which you would like to dump information from one or more source images. If the source images are more extensive than the destination image, i.e. there are areas that extend beyond the destination image boundaries, these areas will not be represented in the updated image. Grid cells in the destination image that are not overlapping with any of the input source images will not be updated, i.e. they will possess the same value as before the resampling operation. The Mosaic tool is used when there is no existing destination image. In this case, a new image is created that represents the bounding rectangle of each of the two or more input images. Grid cells in the output image that do not overlap with any of the input images will be assigned the NoData value.

See Also

mosaic_with_feathering

Function Signature

def mosaic(self, images: List[Raster], resampling_method: str = "cc") -> Raster: ...

mosaic_with_feathering

This tool will create a mosaic from two input images. It is similar in operation to the mosaic tool, however, this tool is the preferred method of mosaicing images when there is significant overlap between the images. For areas of overlap, the feathering method will calculate the output value as a weighted combination of the two input values, where the weights are derived from the squared distance of the pixel to the edge of the data in each of the input raster files. Therefore, less weight is assigned to an image's pixel value where the pixel is very near the edge of the image. Note that the distance is actually calculated to the edge of the grid and not necessarily the edge of the data, which can differ if the image has been rotated during registration. The result of this feathering method is that the output mosaic image should have very little evidence of the original image edges within the overlapping area.

Unlike the Mosaic tool, which can take multiple input images, this tool only accepts two input images. Mosaic is therefore useful when there are many, adjacent or only slightly overlapping images, e.g. for tiled data sets.

Users may want to use the histogram_matching tool prior to mosaicing if the two input images differ significantly in their radiometric properties. i.e. if image contrast differences exist.

See Also

mosaic, histogram_matching

Function Signature

def mosaic_with_feathering(self, image1: Raster, image2: Raster, resampling_method: str = "cc", distance_weight: float = 4.0) -> Raster: ...

multidirectional_hillshade

This tool performs a hillshade operation (also called shaded relief) on an input digital elevation model (DEM) with multiple sources of illumination. The user must input a DEM (dem). Other parameters that must be specified include the altitude of the illumination sources (altitude; i.e. the elevation of the sun above the horizon, measured as an angle from 0 to 90 degrees) and the Z conversion factor (zfactor). The Z conversion factor is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z conversion factor.

The hillshade value (HS) of a DEM grid cell is calculate as:

HS = tan(s) / [1 - tan(s)2]0.5 x [sin(Alt) / tan(s) - cos(Alt) x sin(Az - a)]

where s and a are the local slope gradient and aspect (orientation) respectively and Alt and Az are the illumination source altitude and azimuth respectively. Slope and aspect are calculated using Horn's (1981) 3rd-order finate difference method.

Lastly, the user must specify whether or not to use full 360-degrees of illumination sources (full_mode). When this flag is not specified, the tool will perform a weighted summation of the hillshade images from four illumination azimuth positions at 225, 270, 315, and 360 (0) degrees, given weights of 0.1, 0.4, 0.4, and 0.1 respectively. When run in the full 360-degree mode, eight illumination source azimuths are used to calculate the output at 0, 45, 90, 135, 180, 225, 270, and 315 degrees, with weights of 0.15, 0.125, 0.1, 0.05, 0.1, 0.125, 0.15, and 0.2 respectively.

Classic hillshade (Azimuth=315, Altitude=45.0)

Multi-directional hillshade (Altitude=45.0, Four-direction mode)

Multi-directional hillshade (Altitude=45.0, 360-degree mode)

See Also

hillshade, hypsometrically_tinted_hillshade, aspect, slope

Function Signature

def multidirectional_hillshade(self, dem: Raster, altitude: float = 30.0, z_factor: float = 1.0, full_360_mode: bool = False) -> Raster: ...

multipart_to_singlepart

This tool can be used to convert a vector file containing multi-part features into a vector containing only single-part features. Any multi-part polygons or lines within the input vector file will be split into separate features in the output file, each possessing their own entry in the associated attribute file. For polygon-type vectors, the user may optionally choose to exclude hole-parts from being separated from their containing polygons. That is, with the exclude_holes parameter, hole parts in the input vector will continue to belong to their enclosing polygon in the output vector. The tool will also convert MultiPoint Shapefiles into single Point vectors.

See Also

single_part_to_multipart

Function Signature

def multipart_to_singlepart(self, input: Vector, exclude_holes: bool = False) -> Vector: ...

multiply_overlay

This tool multiplies a stack of raster images (inputs) on a pixel-by-pixel basis. This tool is particularly well suited when you need to create a masking layer from the combination of several Boolean rasters, i.e. for constraint mapping applications. NoData values in any of the input images will result in a NoData pixel in the output image (output).

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

sum_overlay, weighted_sum

Function Signature

def multiply_overlay(self, input_rasters: List[Raster]) -> Raster: ...

multiscale_elevation_percentile

This tool calculates the most elevation percentile (EP) across a range of spatial scales. EP is a measure of local topographic position (LTP) and expresses the vertical position for a digital elevation model (DEM) grid cell (z0) as the percentile of the elevation distribution within the filter window, such that:

EP = counti∈C(zi > z0) x (100 / nC)

where z0 is the elevation of the window's center grid cell, zi is the elevation of cell i contained within the neighboring set C, and nC is the number of grid cells contained within the window.

EP is unsigned and expressed as a percentage, bound between 0% and 100%. This tool outputs two rasters, the multiscale EP magnitude (out_mag) and the scale at which the most extreme EP value occurs (out_scale). The magnitude raster is the most extreme EP value (i.e. the furthest from 50%) for each grid cell encountered within the tested scales of EP.

Quantile-based estimates (e.g., the median and interquartile range) are often used in nonparametric statistics to provide data variability estimates without assuming the distribution is normal. Thus, EP is largely unaffected by irregularly shaped elevation frequency distributions or by outliers in the DEM, resulting in a highly robust metric of LTP. In fact, elevation distributions within small to medium sized neighborhoods often exhibit skewed, multimodal, and non-Gaussian distributions, where the occurrence of elevation errors can often result in distribution outliers. Thus, based on these statistical characteristics, EP is considered one of the most robust representation of LTP.

The algorithm implemented by this tool uses the relatively efficient running-histogram filtering algorithm of Huang et al. (1979). Because most DEMs contain floating point data, elevation values must be rounded to be binned. The sig_digits parameter is used to determine the level of precision preserved during this binning process. The algorithm is parallelized to further aid with computational efficiency.

Experience with multiscale EP has shown that it is highly variable at shorter scales and changes more gradually at broader scales. Therefore, a nonlinear scale sampling interval is used by this tool to ensure that the scale sampling density is higher for short scale ranges and coarser at longer tested scales, such that:

ri = rL + [step × (i - rL)]p

Where ri is the filter radius for step i and p is the nonlinear scaling factor (step_nonlinearity) and a step size (step) of step.

References

Newman, D. R., Lindsay, J. B., and Cockburn, J. M. H. (2018). Evaluating metrics of local topographic position for multiscale geomorphometric analysis. Geomorphology, 312, 40-50.

Huang, T., Yang, G.J.T.G.Y. and Tang, G., 1979. A fast two-dimensional median filtering algorithm. IEEE Transactions on Acoustics, Speech, and Signal Processing, 27(1), pp.13-18.

See Also

elevation_percentile, max_elevation_deviation, max_difference_from_mean

Function Signature

def multiscale_elevation_percentile(self, dem: Raster, num_significant_digits: int = 3, min_scale: int = 4, step_size: int = 1, num_steps: int = 10, step_nonlinearity: float = 1.0) -> Tuple[Raster, Raster]: ...

multiscale_roughness

/

Function Signature

def multiscale_roughness(self, dem: Raster, min_scale: int = 1, max_scale: int = 100, step_size: int = 1) -> Tuple[Raster, Raster]: ...

multiscale_roughness_signature

/

Function Signature

def multiscale_roughness_signature(self, dem: Raster, points: Vector, output_html_file: str, min_scale: int = 1, max_scale: int = 100, step_size: int = 1) -> None: ...

multiscale_std_dev_normals

This tool can be used to map the spatial pattern of maximum spherical standard deviation (σs max; out_mag), as well as the scale at which maximum spherical standard deviation occurs (rmax; out_scale), for each grid cell in an input DEM (dem). This serves as a multi-scale measure of surface roughness, or topographic complexity. The spherical standard deviation (σs) is a measure of the angular spread among n unit vectors and is defined as:

σs = √[-2ln(R / N)] × 180 / π

Where R is the resultant vector length and is derived from the sum of the x, y, and z components of each of the n normals contained within a filter kernel, which designates a tested spatial scale. Each unit vector is a 3-dimensional measure of the surface orientation and slope at each grid cell center. The maximum spherical standard deviation is:

σs max=max{σs(r):r=rL...rU},

Experience with roughness scale signatures has shown that σs max is highly variable at shorter scales and changes more gradually at broader scales. Therefore, a nonlinear scale sampling interval is used by this tool to ensure that the scale sampling density is higher for short scale ranges and coarser at longer tested scales, such that:

ri = rL + [step × (i - rL)]p

Where ri is the filter radius for step i and p is the nonlinear scaling factor (step_nonlinearity) and a step size (step) of step.

Use the spherical_std_dev_of_normals tool if you need to calculate σs for a single scale.

Reference

JB Lindsay, DR Newman, and A Francioni. 2019 Scale-Optimized Surface Roughness for Topographic Analysis. Geosciences, 9(322) doi: 10.3390/geosciences9070322.

See Also

spherical_std_dev_of_normals, multiscale_std_dev_normals_signature, multiscale_roughness

Function Signature

def multiscale_std_dev_normals(self, dem: Raster, min_scale: int = 4, step_size: int = 1, num_steps: int = 10, step_nonlinearity: float = 1.0, html_signature_file: str = "") -> Tuple[Raster, Raster]: ...

multiscale_std_dev_normals_signature

/

Function Signature

def multiscale_std_dev_normals_signature(self, dem: Raster, points: Vector, output_html_file: str, min_scale: int = 4, step_size: int = 1, num_steps: int = 10, step_nonlinearity: float = 1.0) -> None: ...

multiscale_topographic_position_image

This tool creates a multiscale topographic position (MTP) image (see here for an example) from three DEVmax rasters of differing spatial scale ranges. Specifically, multiscale_topographic_position_image takes three DEVmax magnitude rasters, created using the max_elevation_deviation tool, as inputs. The three inputs should correspond to the elevation deviations in the local (local), meso (meso), and broad (broad) scale ranges and will be forced into the blue, green, and red colour components of the colour composite output (output) raster. The image lightness value (lightness) controls the overall brightness of the output image, as depending on the topography and scale ranges, these images can appear relatively dark. Higher values result in brighter, more colourful output images.

The output images can take some training to interpret correctly and a detailed explanation can be found in Lindsay et al. (2015). Sites within the landscape that occupy prominent topographic positions, either low-lying or elevated, will be apparent by their bright colouring in the MTP image. Those that are coloured more strongly in the blue are promient at the local scale range; locations that are more strongly green coloured are promient at the meso scale; and bright reds in the MTP image are associated with broad-scale landscape prominence. Of course, combination colours are also possible when topography is elevated or low-lying across multiple scale ranges. For example, a yellow area would indicated a site of prominent topographic position across the meso and broadest scale ranges.

Reference

Lindsay J, Cockburn J, Russell H. 2015. An integral image approach to performing multi-scale topographic position analysis. Geomorphology, 245: 51-61.

See Also

max_elevation_deviation

Function Signature

def multiscale_topographic_position_image(self, local: Raster, meso: Raster, broad: Raster, lightness: float = 1.2) -> Raster: ...

narrowness_index

This tools calculates a type of shape narrowness index (NI) for raster objects. The index is equal to:

NI = A / (πMD2)

where A is the patch area and MD is the maximum distance-to-edge of the patch. Circular-shaped patches will have a narrowness index near 1.0, while more narrow patch shapes will have higher index values. The index may be conceptualized as the ratio of the patch area to the area of the largest contained circle, although in practice the circle defined by the radius of the maximum distance-to-edge will often fall outside the patch boundaries.

Objects in the input raster (input) are designated by their unique identifiers. Identifier values must be positive, non-zero whole numbers. It is quite common for identifiers to be set using the clump tool applied to some kind of thresholded raster.

See Also

linearity_index, elongation_ratio, clump

Function Signature

def narrowness_index(self, raster: Raster) -> Raster: ...

natural_neighbour_interpolation

This tool can be used to interpolate a set of input vector points (input) onto a raster grid using Sibson's (1981) natural neighbour method. Similar to inverse-distance-weight interpolation (idw_interpolation), the natural neighbour method performs a weighted averaging of nearby point values to estimate the attribute (field) value at grid cell intersections in the output raster (output). However, the two methods differ quite significantly in the way that neighbours are identified and in the weighting scheme. First, natural neigbhour identifies neighbours to be used in the interpolation of a point by finding the points connected to the estimated value location in a Delaunay triangulation, that is, the so-called natural neighbours. This approach has the main advantage of not having to specify an arbitrary search distance or minimum number of nearest neighbours like many other interpolators do. Weights in the natural neighbour scheme are determined using an area-stealing approach, whereby the weight assigned to a neighbour's value is determined by the proportion of its Voronoi polygon that would be lost by inserting the interpolation point into the Voronoi diagram. That is, inserting the interpolation point into the Voronoi diagram results in the creation of a new polygon and shrinking the sizes of the Voronoi polygons associated with each of the natural neighbours. The larger the area by which a neighbours polygon is reduced through the insertion, relative to the polygon of the interpolation point, the greater the weight given to the neighbour point's value in the interpolation. Interpolation weights sum to one because the sum of the reduced polygon areas must account for the entire area of the interpolation points polygon.

The user must specify the attribute field containing point values (field). Alternatively, if the input Shapefile contains z-values, the interpolation may be based on these values (use_z). Either an output grid resolution (cell_size) must be specified or alternatively an existing base file (base) can be used to determine the output raster's (output) resolution and spatial extent. Natural neighbour interpolation generally produces a satisfactorily smooth surface within the region of data points but can produce spurious breaks in the surface outside of this region. Thus, it is recommended that the output surface be clipped to the convex hull of the input points (clip).

Reference

Sibson, R. (1981). "A brief description of natural neighbor interpolation (Chapter 2)". In V. Barnett (ed.). Interpolating Multivariate Data. Chichester: John Wiley. pp. 21–36.

See Also

idw_interpolation, NearestNeighbourGridding

Function Signature

def natural_neighbour_interpolation(self, points: Vector, field_name: str = "FID", use_z: bool = False, cell_size: float = 0.0, base_raster: Raster = None, clip_to_hull: bool = True) -> Raster: ...

nearest_neighbour_interpolation

Creates a raster grid based on a set of vector points and assigns grid values using the nearest neighbour.

Function Signature

def nearest_neighbour_interpolation(self, points: Vector, field_name: str = "FID", use_z: bool = False, cell_size: float = 0.0, base_raster: Raster = None, max_dist: float = float('inf')) -> Raster: ...

new_lidar

Creates a new Lidar object using an input LidarHeader.

Parameters

  • header: LidarHeader - a Lidar header object.

new_raster

Creates a new in-memory Raster object based on a RasterConfig.

Parameters

  • configs: RasterConfigs - An in-memory raster configs object. This can be copied from an existing file, or created manually.

new_raster_from_base_raster

This tool can be used to create a new raster with the same coordinates and dimensions (i.e. rows and columns) as an existing base image. The user must input a base file (base), the value that the new grid will be filled with (out_val; NoData if unspecified), and the data type (data_type flag; options include 'double', 'float', and 'integer').

See Also

new_raster_from_base_vector, raster_cell_assignment

Function Signature

def new_raster_from_base_raster(self, base: Raster, out_val: float = float('nan'), data_type: str = "float") -> Raster: ...

new_raster_from_base_vector

This tool can be used to create a new raster with the same spatial extent as an input vector file (base). The user must specify the name of the base file, the value that the new grid will be filled with (out_val; NoData if unspecified), and the data type (data_type flag; options include 'double', 'float', and 'integer'). It is also necessary to specify a value for the optional grid cell size (cell_size) input parameter.

See Also

new_raster_from_base_raster, raster_cell_assignment

Function Signature

def new_raster_from_base_vector(self, base: Vector, cell_size: float, out_val: float = float('nan'), data_type: str = "float") -> Raster: ...

new_vector

Creates a new in-memory Vector object.

Parameters

  • vector_type: VectorGeometryType - Determines what type of vector data this object can hold. Backed by the Shapefile, the Vector is limited to a single VectorGeometryType.
  • attributes: List[AttributeField] - A list containing the attributes held within the attribute table. Default is None.
  • proj: str - The projection string to be written to the associated *.prj file when written to disc. Default is the empty string.

normal_vectors

Calculates normal vectors for points within a LAS file and stores these data (XYZ vector components) in the RGB field.

Function Signature

def normal_vectors(self, input: Lidar, search_radius: float = -1.0) -> Lidar: ...

normalized_difference_index

This tool can be used to calculate a normalized difference index (NDI) from two bands of multispectral image data. A NDI of two band images (image1 and image2) takes the general form:

NDI = (image1 - image2) / (image1 + image2 + c)

Where c is a correction factor sometimes used to avoid division by zero. It is, however, often set to 0.0. In fact, the normalized_difference_index tool will set all pixels where image1 + image2 = 0 to 0.0 in the output image. While this is not strictly mathematically correct (0 / 0 = infinity), it is often the intended output in these cases.

NDIs generally takes the value range -1.0 to 1.0, although in practice the range of values for a particular image scene may be more restricted than this.

NDIs have two important properties that make them particularly useful for remote sensing applications. First, they emphasize certain aspects of the shape of the spectral signatures of different land covers. Secondly, they can be used to de-emphasize the effects of variable illumination within a scene. NDIs are therefore frequently used in the field of remote sensing to create vegetation indices and other indices for emphasizing various land-covers and as inputs to analytical operations like image classification. For example, the normalized difference vegetation index (NDVI), one of the most common image-derived products in remote sensing, is calculated as:

NDVI = (NIR - RED) / (NIR + RED)

The optimal soil adjusted vegetation index (OSAVI) is:

OSAVI = (NIR - RED) / (NIR + RED + 0.16)

The normalized difference water index (NDWI), or normalized difference moisture index (NDMI), is:

NDWI = (NIR - SWIR) / (NIR + SWIR)

The normalized burn ratio 1 (NBR1) and normalized burn ration 2 (NBR2) are:

NBR1 = (NIR - SWIR2) / (NIR + SWIR2)

NBR2 = (SWIR1 - SWIR2) / (SWIR1 + SWIR2)

In addition to NDIs, Simple Ratios of image bands, are also commonly used as inputs to other remote sensing applications like image classification. Simple ratios can be calculated using the Divide tool. Division by zero, in this case, will result in an output NoData value.

See Also

Divide

Function Signature

def normalized_difference_index(self, nir_image: Raster, red_image: Raster, clip_percent: float = 0.0, correction_value: float = 0.0) -> Raster: ...

num_downslope_neighbours

This tool calculates the number of downslope neighbours of each grid cell in a raster digital elevation model (DEM). The user must input a DEM (dem). The tool examines the eight neighbouring cells for each grid cell in a the DEM and counts the number of neighbours with an elevation less than the centre cell of the 3 x 3 window. The output image can therefore have values raning from 0 to 8. A raster grid cell with eight downslope neighbours is a peak and a cell with zero downslope neighbours is a pit. This tool can be used with the NumUpslopeNeighbours tool to assess the degree of local flow divergence/convergence.

See Also

NumUpslopeNeighbours

Function Signature

def num_downslope_neighbours(self, dem: Raster) -> Raster: ...

num_inflowing_neighbours

This tool calculates the number of inflowing neighbours for each grid cell in a raster file. The user must specify the names of an input digital elevation model (DEM) file (dem) and the output raster file (output). The tool calculates the D8 pointer file internally in order to identify inflowing neighbouring cells.

Grid cells in the input DEM that contain the NoData value will be assigned the NoData value in the output image. The output image is of the integer data type and continuous data scale.

See Also

num_downslope_neighbours, NumUpslopeNeighbours

Function Signature

def num_inflowing_neighbours(self, dem: Raster) -> Raster: ...

olympic_filter

This filter is a modification of the mean_filter, whereby the highest and lowest values in the kernel are dropped, and the remaining values are averaged to replace the central pixel. The result is a low-pass smoothing filter that is more robust than the mean_filter, which is more strongly impacted by the presence of outlier values. It is named after a system of scoring Olympic events.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

See Also

mean_filter

Function Signature

def olympic_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

opening

This tool performs an opening operation on an input greyscale image (input). An opening is a mathematical morphology operation involving a dilation (maximum filter) on an erosion (minimum filter) set. opening operations, together with the closing operation, is frequently used in the fields of computer vision and digital image processing for image noise removal. The user must specify the size of the moving window in both the x and y directions (filterx and filtery).

See Also

closing, tophat_transform

Function Signature

def opening(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

paired_sample_t_test

This tool will perform a paired-sample t-test to evaluate whether a significant statistical difference exists between the two rasters. The null hypothesis is that the difference between the paired population means is equal to zero. The paired-samples t-test makes an assumption that the differences between related samples follows a Gaussian distribution. The tool will output a cumulative probability distribution, with a fitted Gaussian, to help users evaluate whether this assumption is violated by the data. If this is the case, the wilcoxon_signed_rank_test should be used instead.

The user must specify the name of the two input raster images (input1 and input2) and the output report HTML file (output). The test can be performed optionally on the entire image or on a random sub-sample of pixel values of a user-specified size (num_samples). In evaluating the significance of the test, it is important to keep in mind that given a sufficiently large sample, extremely small and non-notable differences can be found to be statistically significant. Furthermore statistical significance says nothing about the practical significance of a difference.

See Also

two_sample_ks_test, wilcoxon_signed_rank_test

Function Signature

def paired_sample_t_test(self, raster1: Raster, raster2: Raster, output_html_file: str, num_samples: int) -> None: ...

panchromatic_sharpening

Panchromatic sharpening, or simply pan-sharpening, refers to a range of techniques that can be used to merge finer spatial resolution panchromatic images with coarser spatial resolution multi-spectral images. The multi-spectral data provides colour information while the panchromatic image provides improved spatial information. This procedure is sometimes called image fusion. Jensen (2015) describes panchromatic sharpening in detail.

Whitebox provides two common methods for panchromatic sharpening including the Brovey transformation and the Intensity-Hue-Saturation (IHS) methods. Both of these techniques provide the best results when the range of wavelengths detected by the panchromatic image overlap significantly with the wavelength range covered by the three multi-spectral bands that are used. When this is not the case, the resulting colour composite will likely have colour properties that are dissimilar to the colour composite generated by the original multispectral images. For Landsat ETM+ data, the panchromatic band is sensitive to EMR in the range of 0.52-0.90 micrometres. This corresponds closely to the green (band 2), red (band 3), and near-infrared (band 4).

Reference

Jensen, J. R. (2015). Introductory Digital Image Processing: A Remote Sensing Perspective.

See Also

create_colour_composite

Function Signature

def panchromatic_sharpening(self, pan: Raster, colour_composite: Raster, red: Raster, green: Raster, blue: Raster, fusion_method: str = "brovey") -> Raster: ...

patch_orientation

This tool calculates the orientation of polygon features based on the slope of a reduced major axis (RMA) regression line. The regression analysis use the vertices of the exterior hull nodes of a vector polygon. The only required input is the name of the vector polygon file. The orientation values, measured in degrees from north, will be placed in the accompanying attribute table as a new field (ORIENT). The value of the orientation measure for any polygon will depend on how elongated the feature is.

Note that the output values are polygon orientations and not true directions. While directions may take values ranging from 0-360, orientation is expressed as an angle between 0 and 180 degrees clockwise from north. Lastly, the orientation measure may become unstable when polygons are oriented nearly vertical or horizontal.

See Also

linearity_index, elongation_ratio

Function Signature

def patch_orientation(self, input: Vector) -> Vector: ...

pennock_landform_classification

Tool can be used to perform a simple landform classification based on measures of slope gradient and curvature derived from a user-specified digital elevation model (DEM). The classification scheme is based on the method proposed by Pennock, Zebarth, and DeJong (1987). The scheme divides a landscape into seven element types, including: convergent footslopes (CFS), divergent footslopes (DFS), convergent shoulders (CSH), divergent shoulders (DSH), convergent backslopes (CBS), divergent backslopes (DBS), and level terrain (L). The output raster image will record each of these base element types as:

Element TypeCode
CFS1
DFS2
CSH3
DSH4
CBS5
DBS6
L7

The definition of each of the elements, based on the original Pennock et al. (1987) paper, is as follows:

PROFILEGRADIENTPLANElement
Concave ( -0.10)High >3.0Concave 0.0CFS
Concave ( -0.10)High >3.0Convex >0.0DFS
Convex (>0.10)High >3.0Concave 0.0CSH
Convex (>0.10)High >3.0Convex >0.0DSH
Linear (-0.10...0.10)High >3.0Concave 0.0CBS
Linear (-0.10...0.10)High >3.0Convex >0.0DBS
--Low 3.0--L

Where PROFILE is profile curvature, GRADIENT is the slope gradient, and PLAN is the plan curvature. Note that these values are likely landscape and data specific and can be adjusted by the user. Landscape classification schemes that are based on terrain attributes are highly sensitive to short-range topographic variability (i.e. roughness) and can benefit from pre-processing the DEM with a smoothing filter to reduce the effect of surface roughness and emphasize the longer-range topographic signal. The feature_preserving_smoothing tool offers excellent performance in smoothing DEMs without removing the sharpness of breaks-in-slope.

Reference

Pennock, D.J., Zebarth, B.J., and DeJong, E. (1987) Landform classification and soil distribution in hummocky terrain, Saskatchewan, Canada. Geoderma, 40: 297-315.

See Also

feature_preserving_smoothing

Function Signature

def pennock_landform_classification(self, dem: Raster, slope_threshold: float = 3.0, prof_curv_threshold: float = 0.1, plan_curv_threshold: float = 0.0, z_factor: float = 1.0) -> Tuple[Raster, str]: ...

percent_elev_range

Percent elevation range (PER) is a measure of local topographic position (LTP). It expresses the vertical position for a digital elevation model (DEM) grid cell (z0) as the percentage of the elevation range within the neighbourhood filter window, such that:

PER = z0 / (zmax - zmin) x 100

where z0 is the elevation of the window's center grid cell, zmax is the maximum neighbouring elevation, and zmin is the minimum neighbouring elevation.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filteryflags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

Compared with ElevPercentile and DevFromMeanElev, PER is a less robust measure of LTP that is susceptible to outliers in neighbouring elevations (e.g. the presence of off-terrain objects in the DEM).

References

Newman, D. R., Lindsay, J. B., and Cockburn, J. M. H. (2018). Evaluating metrics of local topographic position for multiscale geomorphometric analysis. Geomorphology, 312, 40-50.

See Also

ElevPercentile, DevFromMeanElev, DiffFromMeanElev, relative_topographic_position

Function Signature

def percent_elev_range(self, dem: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

percent_equal_to

This tool calculates the percentage of a raster stack (inputs) that have cell values equal to an input comparison raster. The user must specify the name of the value raster (comparison), the names of the raster files contained in the stack, and an output raster file name (output). The tool, working on a cell-by-cell basis, will count the number of rasters within the stack that have the same grid cell value as the corresponding grid cell in the comparison raster. This count is then expressed as a percentage of the number of rasters contained within the stack and output. If any of the rasters within the stack contain the NoData value, the corresponding grid cell in the output raster will be assigned NoData.

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

percent_greater_than, percent_less_than

Function Signature

def percent_equal_to(self, input_rasters: List[Raster], comparison: Raster) -> Raster: ...

percent_greater_than

This tool calculates the percentage of a raster stack (inputs) that have cell values greater than an input comparison raster. The user must specify the name of the value raster (comparison), the names of the raster files contained in the stack, and an output raster file name (output). The tool, working on a cell-by-cell basis, will count the number of rasters within the stack with larger grid cell values greater than the corresponding grid cell in the comparison raster. This count is then expressed as a percentage of the number of rasters contained within the stack and output. If any of the rasters within the stack contain the NoData value, the corresponding grid cell in the output raster will be assigned NoData.

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

percent_less_than, percent_equal_to

Function Signature

def percent_greater_than(self, input_rasters: List[Raster], comparison: Raster) -> Raster: ...

percent_less_than

This tool calculates the percentage of a raster stack (inputs) that have cell values less than an input comparison raster. The user must specify the name of the value raster (comparison), the names of the raster files contained in the stack, and an output raster file name (output). The tool, working on a cell-by-cell basis, will count the number of rasters within the stack with larger grid cell values less than the corresponding grid cell in the comparison raster. This count is then expressed as a percentage of the number of rasters contained within the stack and output. If any of the rasters within the stack contain the NoData value, the corresponding grid cell in the output raster will be assigned NoData.

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

percent_greater_than, percent_equal_to

Function Signature

def percent_less_than(self, input_rasters: List[Raster], comparison: Raster) -> Raster: ...

percentage_contrast_stretch

This tool performs a percentage contrast stretch on a raster image. This operation maps each grid cell value in the input raster image (zin) onto a new scale that ranges from a lower-tail clip value (min_val) to the upper-tail clip value (max_val), with the user-specified number of tonal values (num_tones), such that:

zout = ((zin – min_val)/(max_val – min_val)) x num_tones

where zout is the output value. The values of min_val and max_val are determined from the frequency distribution and the user-specified tail clip value (clip). For example, if a value of 1% is specified, the tool will determine the values in the input image for which 1% of the grid cells have a lower value min_val and 1% of the grid cells have a higher value max_val. The user must also specify which tails (upper, lower, or both) to clip (tail).

This is a type of linear contrast stretch with saturation at the tails of the frequency distribution. This is the same kind of stretch that is used to display raster type data on the fly in many GIS software packages, such that the lower and upper tail values are set using the minimum and maximum display values and the number of tonal values is determined by the number of palette entries.

See Also

PiecewiseContrastStretch, gaussian_contrast_stretch, histogram_equalization, min_max_contrast_stretch, sigmoidal_contrast_stretch, standard_deviation_contrast_stretch

Function Signature

def percentage_contrast_stretch(self, raster: Raster, clip: float = 1.0, tail: str = "both", num_tones: int = 256) -> Raster: ...

percentile_filter

This tool calculates the percentile of the center cell in a moving filter window applied to an input image (`input). This indicates the value below which a given percentage of the neighbouring values in within the filter fall. For example, the 35th percentile is the value below which 35% of the neighbouring values in the filter window may be found. As such, the percentile of a pixel value is indicative of the relative location of the site within the statistical distribution of values contained within a filter window. When applied to input digital elevation models, percentile is a measure of local topographic position, or elevation residual.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values, e.g. 3, 5, 7, 9... If the kernel filter size is the same in the x and y dimensions, the silent filter flag may be used instead (command-line interface only).

This tool takes advantage of the redundancy between overlapping, neighbouring filters to enhance computationally efficiency, using a method similar to Huang et al. (1979). This efficient method of calculating percentiles requires rounding of floating-point inputs, and therefore the user must specify the number of significant digits (sig_digits) to be used during the processing. Like most of WhiteboxTools' filters, this tool is also parallelized for further efficiency.

Reference

Huang, T., Yang, G.J.T.G.Y. and Tang, G., 1979. A fast two-dimensional median filtering algorithm. IEEE Transactions on Acoustics, Speech, and Signal Processing, 27(1), pp.13-18.

See Also

median_filter

Function Signature

def percentile_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11, sig_digits: int = 2) -> Raster: ...

perimeter_area_ratio

The perimeter-area ratio is an indicator of polygon shape complexity. Unlike some other shape parameters (e.g. shape complexity index), perimeter-area ratio does not standardize to a simple Euclidean shape. Although widely used for landscape analysis, perimeter-area ratio exhibits the undesirable property of polygon size dependence (Mcgarigal et al. 2002). That is, holding shape constant, an increase in polygon size will cause a decrease in the perimeter-area ratio. The perimeter-area ratio is the inverse of the compactness ratio.

The output data will be displayed as a new field (P_A_RATIO) in the input vector's database file.

Function Signature

def perimeter_area_ratio(self, input: Vector) -> Vector: ...

pick_from_list

This tool outputs the cell value from a raster stack specified (inputs) by a position raster (pos_input). The user must specify the name of the position raster, the names of the raster files contained in the stack (i.e. group of rasters), and an output raster file name (output). The tool, working on a cell-by-cell basis, will assign the value to the output grid cell contained in the corresponding cell in the stack image in the position specified by the cell value in the position raster. Importantly, the positions raster should be in zero-based order. That is, the first image in the stack should be assigned the value zero, the second raster is assigned 1, and so on.

At least two input rasters are required to run this tool. Each of the input rasters must share the same number of rows and columns and spatial extent. An error will be issued if this is not the case.

See Also

count_if

Function Signature

def pick_from_list(self, input_rasters: List[Raster], pos_input: Raster) -> Raster: ...

plan_curvature

This tool calculates the plan curvature (i.e. contour curvature), or the rate of change in aspect along a contour line, from a digital elevation model (DEM). Curvature is the second derivative of the topographic surface defined by a DEM. Plan curvature characterizes the degree of flow convergence or divergence within the landscape (Gallant and Wilson, 2000). The user must input a DEM (dem). WhiteboxTools reports curvature in degrees multiplied by 100 for easier interpretation. The Z conversion factor (zfactor) is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z Conversion Factor. If the DEM is in the geographic coordinate system (latitude and longitude), the following equation is used:

zfactor = 1.0 / (111320.0 x cos(mid_lat))

where mid_lat is the latitude of the centre of the raster, in radians.

The algorithm uses the same formula for the calculation of plan curvature as Gallant and Wilson (2000). Plan curvature is negative for diverging flow along ridges and positive for convergent areas, e.g. along valley bottoms.

Reference

Gallant, J. C., and J. P. Wilson, 2000, Primary topographic attributes, in Terrain Analysis: Principles and Applications, edited by J. P. Wilson and J. C. Gallant pp. 51-86, John Wiley, Hoboken, N.J.

See Also

profile_curvature, tangential_curvature, total_curvature, slope, aspect

Function Signature

def plan_curvature(self, dem: Raster, log_transform: bool = False, z_factor: float = 1.0) -> Raster: ...

polygon_area

This tool calculates the area of vector polygons, adding the result to the vector's attribute table (AREA field). The area calculation will account for any holes contained within polygons. The vector should be in a projected coordinate system.

To calculate the area of raster polygons, use the raster_area tool instead.

See Also

raster_area

Function Signature

def polygon_area(self, input: Vector) -> Vector: ...

polygon_long_axis

This tool can be used to map the long axis of polygon features. The long axis is the longer of the two primary axes of the minimum bounding box (MBB), i.e. the smallest box to completely enclose a feature. The long axis is drawn for each polygon in the input vector file such that it passes through the centre point of the MBB. The output file is therefore a vector of simple two-point polylines forming a vector field.

Function Signature

def polygon_long_axis(self, input: Vector) -> Vector: ...

polygon_perimeter

This tool calculates the perimeter of vector polygons, adding the result to the vector's attribute table (PERIMETER field). The area calculation will account for any holes contained within polygons. The vector should be in a a projected coordinate system.

Function Signature

def polygon_perimeter(self, input: Vector) -> Vector: ...

polygon_short_axis

This tool can be used to map the short axis of polygon features. The short axis is the shorter of the two primary axes of the minimum bounding box (MBB), i.e. the smallest box to completely enclose a feature. The short axis is drawn for each polygon in the input vector file such that it passes through the centre point of the MBB. The output file is therefore a vector of simple two-point polylines forming a vector field.

Function Signature

def polygon_short_axis(self, input: Vector) -> Vector: ...

polygonize

This tool outputs a vector polygon layer from two or more intersecting line features contained in one or more input vector line files. Each space enclosed by the intersecting line set is converted to polygon added to the output layer. This tool should not be confused with the lines_to_polygons tool, which can be used to convert a vector file of polylines into a set of polygons, simply by closing each line feature. The lines_to_polygons tool does not deal with line intersection in the same way that the polygonize tool does.

See Also

lines_to_polygons

Function Signature

def polygonize(self, input_layers: List[Vector]) -> Vector: ...

polygons_to_lines

This tool converts vector polygons into polylines, simply by modifying the Shapefile geometry type.

See Also

lines_to_polygons

Function Signature

def polygons_to_lines(self, input: Vector) -> Vector: ...

prewitt_filter

This tool performs a 3 × 3 Prewitt edge-detection filter on a raster image. The Prewitt filter is similar to the sobel_filter, in that it identifies areas of high slope in the input image through the calculation of slopes in the x and y directions. The Prewitt edge-detection filter, however, gives less weight to nearer cell values within the moving window, or kernel. For example, a Prewitt filter uses the following schemes to calculate x and y slopes:

X-direction slope

...
-101
-101
-101

Y-direction slope

...
111
000
-1-1-1

Each grid cell in the output image is assigned the square-root of the squared sum of the x and y slopes.

The user may optionally clip the output image distribution tails by a specified amount (e.g. 1%).

See Also

sobel_filter

Function Signature

def prewitt_filter(self, raster: Raster, clip_tails: float = 0.0) -> Raster: ...

principal_component_analysis

Principal component analysis (PCA) is a common data reduction technique that is used to reduce the dimensionality of multi-dimensional space. In the field of remote sensing, PCA is often used to reduce the number of bands of multi-spectral, or hyper-spectral, imagery. Image correlation analysis often reveals a substantial level of correlation among bands of multi-spectral imagery. This correlation represents data redundancy, i.e. fewer images than the number of bands are required to represent the same information, where the information is related to variation within the imagery. PCA transforms the original data set of n bands into n 'component' images, where each component image is uncorrelated with all other components. The technique works by transforming the axes of the multi-spectral space such that it coincides with the directions of greatest correlation. Each of these new axes are orthogonal to one another, i.e. they are at right angles. PCA is therefore a type of coordinate system transformation. The PCA component images are arranged such that the greatest amount of variance (or information) within the original data set, is contained within the first component and the amount of variance decreases with each component. It is often the case that the majority of the information contained in a multi-spectral data set can be represented by the first three or four PCA components. The higher-order components are often associated with noise in the original data set.

The user must specify the names of the multiple input images (inputs). Additionally, the user must specify whether to perform a standardized PCA (standardized) and the number of output components (num_comp) to generate (all components will be output unless otherwise specified). A standardized PCA is performed using the correlation matrix rather than the variance-covariance matrix. This is appropriate when the variances in the input images differ substantially, such as would be the case if they contained values that were recorded in different units (e.g. feet and meters) or on different scales (e.g. 8-bit vs. 16 bit).

Several outputs will be generated when the tool has completed. The PCA report will be embedded within an output (output) HTML file, which should be automatically displayed after the tool has completed. This report contains useful data summarizing the results of the PCA, including the explained variances of each factor, the Eigenvalues and Eigenvectors associated with factors, the factor loadings, and a scree plot. The first table that is in the PCA report lists the amount of explained variance (in non-cumulative and cumulative form), the Eigenvalue, and the Eigenvector for each component. Each of the PCA components refer to the newly created, transformed images that are created by running the tool. The amount of explained variance associated with each component can be thought of as a measure of how much information content within the original multi-spectral data set that a component has. The higher this value is, the more important the component is. This same information is presented in graphical form in the scree plot, found at the bottom of the PCA report. The Eigenvalue is another measure of the information content of a component and the eigenvector describes the mathematical transformation (rotation coordinates) that correspond to a particular component image.

Factor loadings are also output in a table within the PCA text report (second table). These loading values describe the correlation (i.e. r values) between each of the PCA components (columns) and the original images (rows). These values show you how the information contained in an image is spread among the components. An analysis of factor loadings can be reveal useful information about the data set. For example, it can help to identify groups of similar images.

PCA is used to reduce the number of band images necessary for classification (i.e. as a data reduction technique), for noise reduction, and for change detection applications. When used as a change detection technique, the major PCA components tend to be associated with stable elements of the data set while variance due to land-cover change tend to manifest in the high-order, 'change components'. When used as a noise reduction technique, an inverse PCA is generally performed, leaving out one or more of the high-order PCA components, which account for noise variance.

Note: the current implementation reads every raster into memory at one time. This is because of the calculation of the co-variances. As such, if the entire image stack cannot fit in memory, the tool will likely experience an out-of-memory error. This tool should be run using the wd flag to specify the working directory into which the component images will be written.

Function Signature

def principal_component_analysis(self, rasters: List[Raster], output_html_file: str, num_components: int = 2, standardized: bool = False) -> List[Raster]: ...

print_geotiff_tags

This tool can be used to view the tags contained within a GeoTiff file. Viewing the tags of a GeoTiff file can be useful when trying to import the GeoTiff to different software environments. The user must specify the name of a GeoTiff file and the tag information will be output to the StdOut output stream (e.g. console). Note that tags that contain greater than 100 values will be truncated in the output. GeoKeys will also be interpreted as per the GeoTIFF specification.

Function Signature

def print_geotiff_tags(self, file_name: str) : ...

profile

This tool can be used to plot the data profile, along a set of one or more vector lines (lines), in an input (surface) digital elevation model (DEM), or other surface model. The data profile plots surface height (y-axis) against distance along profile (x-axis). The tool outputs an interactive SVG line graph embedded in an HTML document (output). If the vector lines file contains multiple line features, the output plot will contain each of the input profiles.

If you want to extract the longitudinal profile of a river, use the long_profile tool instead.

See Also

long_profile, hypsometric_analysis

Function Signature

def profile(self, lines_vector: Vector, surface: Raster, output_html_file: str) -> None: ...

profile_curvature

This tool calculates the profile curvature, or the rate of change in slope along a flow line, from a digital elevation model (DEM). Curvature is the second derivative of the topographic surface defined by a DEM. Profile curvature characterizes the degree of downslope acceleration or deceleration within the landscape (Gallant and Wilson, 2000). The user must input DEM a (dem). WhiteboxTools reports curvature in degrees multiplied by 100 for easier interpretation because curvature values are typically very small. The Z conversion factor (zfactor) is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z Conversion Factor. If the DEM is in the geographic coordinate system (latitude and longitude), the following equation is used:

zfactor = 1.0 / (111320.0 x cos(mid_lat))

where mid_lat is the latitude of the centre of the raster, in radians.

The algorithm uses the same formula for the calculation of plan curvature as Gallant and Wilson (2000). Profile curvature is negative for slope increasing downhill (convex flow profile, typical of upper slopes) and positive for slope decreasing downhill (concave, typical of lower slopes).

Reference

Gallant, J. C., and J. P. Wilson, 2000, Primary topographic attributes, in Terrain Analysis: Principles and Applications, edited by J. P. Wilson and J. C. Gallant pp. 51-86, John Wiley, Hoboken, N.J.

See Also

profile_curvature, tangential_curvature, total_curvature, slope, aspect

Function Signature

def profile_curvature(self, dem: Raster, log_transform: bool = False, z_factor: float = 1.0) -> Raster: ...

qin_flow_accumulation

This tool is used to generate a flow accumulation grid (i.e. contributing area) using the Qin et al. (2007) flow algorithm, not to be confused with the similarly named quinn_flow_accumulation tool. This algorithm is an examples of a multiple-flow-direction (MFD) method because the flow entering each grid cell is routed to more than one downslope neighbour, i.e. flow divergence is permitted. It is based on a modification of the Freeman (1991; FD8FlowAccumulation) and Quinn et al. (1995; quinn_flow_accumulation) methods. The Qin method relates the degree of flow dispersion from a grid cell to the local maximum downslope gradient. Specifically, steeper terrain experiences more convergent flow while flatter slopes experience more flow divergence.

The following equations are used to calculate the portion flow (Fi) given to each neighbour, i:

Fi = Li(tanβ)f(e) / Σi=1n[Li(tanβ)f(e)]

f(e) = min(e, eU) / eU × (pU - 1.1) + 1.1

Where Li is the contour length, and is 0.5×cell size for cardinal directions and 0.354×cell size for diagonal directions, n = 8, and represents each of the eight neighbouring grid cells. The exponent f(e) controls the proportion of flow allocated to each downslope neighbour of a grid cell, based on the local maximum downslope gradient (e), and the user-specified upper boundary of e (eU; max_slope), and the upper boundary of the exponent (pU; exponent), f(e). Note that the original Qin (2007) implementation allowed for user-specified lower boundaries on the slope (eL) and exponent (pL) parameters as well. In this implementation, these parameters are assumed to be 0.0 and 1.1 respectively, and are not user adjustable. Also note, the exponent parameter should be less than 50.0, as higher values may cause numerical instability.

The user must specify the name (dem) of the input digital elevation model (DEM) and the output file (output). The DEM must have been hydrologically corrected to remove all spurious depressions and flat areas. DEM pre-processing is usually achieved using either the breach_depressions_least_cost (also breach_depressions_least_cost) or fill_depressions tool.

The user-specified non-dispersive, channel initiation threshold (threshold) is a flow-accumulation value (measured in upslope grid cells, which is directly proportional to area) above which flow dispersion is no longer permitted. Grid cells with flow-accumulation values above this area threshold will have their flow routed in a manner that is similar to the D8 single-flow-direction algorithm, directing all flow towards the steepest downslope neighbour. This is usually done under the assumption that flow dispersion, whilst appropriate on hillslope areas, is not realistic once flow becomes channelized. Importantly, the threshold parameter sets the spatial extent of the stream network, with lower values resulting in more extensive networks.

In addition to the input DEM, output file (output), and exponent, the user must also specify the output type (out_type). The output flow-accumulation can be: 1) cells (i.e. the number of inflowing grid cells), catchment area (i.e. the upslope area), or specific contributing area (i.e. the catchment area divided by the flow width). The default value is specific contributing area. The user must also specify whether the output flow-accumulation grid should be log-tranformed (log), i.e. the output, if this option is selected, will be the natural-logarithm of the accumulated flow value. This is a transformation that is often performed to better visualize the contributing area distribution. Because contributing areas tend to be very high along valley bottoms and relatively low on hillslopes, when a flow-accumulation image is displayed, the distribution of values on hillslopes tends to be 'washed out' because the palette is stretched out to represent the highest values. Log-transformation provides a means of compensating for this phenomenon. Importantly, however, log-transformed flow-accumulation grids must not be used to estimate other secondary terrain indices, such as the wetness index (wetness_index), or relative stream power index (StreamPowerIndex).

Reference

Freeman, T. G. (1991). Calculating catchment area with divergent flow based on a regular grid. Computers and Geosciences, 17(3), 413-422.

Qin, C., Zhu, A. X., Pei, T., Li, B., Zhou, C., & Yang, L. 2007. An adaptive approach to selecting a flow‐partition exponent for a multiple‐flow‐direction algorithm. International Journal of Geographical Information Science, 21(4), 443-458.

Quinn, P. F., K. J. Beven, Lamb, R. 1995. The in (a/tanβ) index: How to calculate it and how to use it within the topmodel framework. Hydrological Processes 9(2): 161-182.

See Also

D8FlowAccumulation, quinn_flow_accumulation, FD8FlowAccumulation, DInfFlowAccumulation, MDInfFlowAccumulation, rho8_pointer, wetness_index

Function Signature

def qin_flow_accumulation(self, dem: Raster, out_type: str = "sca", exponent: float = 10.0, max_slope: float = 45.0, convergence_threshold: float = float('inf'), log_transform: bool = False, clip: bool = False) -> Raster: ...

quantiles

This tool transforms values in an input raster (input) into quantiles. In statistics, quantiles are cut points dividing the range of a probability distribution into continuous intervals with equal probabilities, or dividing the observations in a sample in a same way. There is one fewer quantile than the number of groups created. Thus quartiles are the three cut points that will divide a dataset into four equal-sized groups. Common quantiles have special names: for instance quartile (4-quantile), quintiles (5-quantiles), decile (10-quantile), percentile (100-quantile).

The user must specify the desired number of quantiles, q (num_quantiles), in the output raster (output). The output raster will contain q equal-sized groups with values 1 to q, indicating which quantile group each grid cell belongs to.

See Also

histogram_equalization

Function Signature

def quantiles(self, raster: Raster, num_quantiles: int = 5) -> Raster: ...

quinn_flow_accumulation

This tool is used to generate a flow accumulation grid (i.e. contributing area) using the Quinn et al. (1995) flow algorithm, sometimes called QMFD or QMFD2, and not to be confused with the similarly named qin_flow_accumulation tool. This algorithm is an examples of a multiple-flow-direction (MFD) method because the flow entering each grid cell is routed to more than one downslope neighbour, i.e. flow divergence is permitted. The user must specify the name (dem) of the input digital elevation model (DEM). The DEM must have been hydrologically corrected to remove all spurious depressions and flat areas. DEM pre-processing is usually achieved using either the breach_depressions_least_cost (also breach_depressions_least_cost) or fill_depressions tool. A value must also be specified for the exponent parameter (exponent), a number that controls the degree of dispersion in the resulting flow-accumulation grid. A lower value yields greater apparent flow dispersion across divergent hillslopes. The exponent value (h) should probably be less than 50.0, as higher values may cause numerical instability, and values between 1 and 2 are most common. The following equations are used to calculate the portion flow (Fi) given to each neighbour, i:

Fi = Li(tanβ)p / Σi=1n[Li(tanβ)p]

p = (A / threshold + 1)h

Where Li is the contour length, and is 0.5×cell size for cardinal directions and 0.354×cell size for diagonal directions, n = 8, and represents each of the eight neighbouring grid cells, and, A is the flow accumulation value assigned to the current grid cell, that is being apportioned downslope. The non-dispersive, channel initiation threshold (threshold) is a flow-accumulation value (measured in upslope grid cells, which is directly proportional to area) above which flow dispersion is no longer permitted. Grid cells with flow-accumulation values above this threshold will have their flow routed in a manner that is similar to the D8 single-flow-direction algorithm, directing all flow towards the steepest downslope neighbour. This is usually done under the assumption that flow dispersion, whilst appropriate on hillslope areas, is not realistic once flow becomes channelized. Importantly, the threshold parameter sets the spatial extent of the stream network, with lower values resulting in more extensive networks.

In addition to the input DEM, output file (output), and exponent, the user must also specify the output type (out_type). The output flow-accumulation can be: 1) cells (i.e. the number of inflowing grid cells), catchment area (i.e. the upslope area), or specific contributing area (i.e. the catchment area divided by the flow width). The default value is specific contributing area. The user must also specify whether the output flow-accumulation grid should be log-transformed (log), i.e. the output, if this option is selected, will be the natural-logarithm of the accumulated flow value. This is a transformation that is often performed to better visualize the contributing area distribution. Because contributing areas tend to be very high along valley bottoms and relatively low on hillslopes, when a flow-accumulation image is displayed, the distribution of values on hillslopes tends to be 'washed out' because the palette is stretched out to represent the highest values. Log-transformation provides a means of compensating for this phenomenon. Importantly, however, log-transformed flow-accumulation grids must not be used to estimate other secondary terrain indices, such as the wetness index (wetness_index), or relative stream power index (StreamPowerIndex). The Quinn et al. (1995) algorithm is commonly used to calculate wetness index.

Reference

Quinn, P. F., K. J. Beven, Lamb, R. 1995. The in (a/tanβ) index: How to calculate it and how to use it within the topmodel framework. Hydrological Processes 9(2): 161-182.

See Also

D8FlowAccumulation, qin_flow_accumulation, FD8FlowAccumulation, DInfFlowAccumulation, MDInfFlowAccumulation, rho8_pointer, wetness_index

Function Signature

def quinn_flow_accumulation(self, dem: Raster, out_type: str = "sca", exponent: float = 1.1, convergence_threshold: float = float('inf'), log_transform: bool = False, clip: bool = False) -> Raster: ...

radial_basis_function_interpolation

This tool interpolates vector points into a raster surface using a radial basis function (RBF) scheme.

Function Signature

def radial_basis_function_interpolation(self, points: Vector, field_name: str = "FID", use_z: bool = False, radius: float = 0.0, min_points: int = 0, cell_size: float = 0.0, base_raster: Raster = None, func_type: str = "thinplatespline", poly_order: str = "none", weight: float = 0.1) -> Raster: ...

radius_of_gyration

This can be used to calculate the radius of gyration (RoG) for the polygon features within a raster image. RoG measures how far across the landscape a polygon extends its reach on average, given by the mean distance between cells in a patch (Mcgarigal et al. 2002). The radius of gyration can be considered a measure of the average distance an organism can move within a patch before encountering the patch boundary from a random starting point (Mcgarigal et al. 2002). The input raster grid should contain polygons with unique identifiers greater than zero. The user must also specify the name of the output raster file (where the radius of gyration will be assigned to each feature in the input file) and the specified option of outputting text data.

Function Signature

def radius_of_gyration(self, raster: Raster) -> Tuple[Raster, str]: ...

raise_walls

This tool is used to increment the elevations in a digital elevation model (DEM) along the boundaries of a vector lines or polygon layer. The user must specify the name of the raster DEM (dem), the vector file (input), the output file name (output), the increment height (height), and an optional breach lines vector layer (breach). The breach lines layer can be used to breach a whole in the raised walls at intersections with the wall layer.

Function Signature

def raise_walls(self, dem: Raster, walls: Vector, breach_lines: Vector, wall_height: float = 100.0) -> Raster: ...

random_field

This tool can be used to a raster image filled with random values drawn from a standard normal distribution. The values range from approximately -4.0 to 4.0, with a mean of 0 and a standard deviation of 1.0. The dimensions and georeferencing of the output random field (output) are based on an existing, user-specified raster grid (base). Note that the output field will not possess any spatial autocorrelation. If spatially autocorrelated random fields are desired, the turning_bands_simulation tool is more appropriate, or alternatively, the fast_almost_gaussian_filter tool may be used to force spatial autocorrelation onto the distribution of the random_field tool.

See Also

turning_bands_simulation, fast_almost_gaussian_filter

Function Signature

def random_field(self, base_raster: Raster = None) -> Raster: ...

random_sample

This tool can be used to create a random sample of grid cells. The user specifies the base raster file, which is used to determine the grid dimensions and georeference information for the output raster, and the number of sample random samples (n). The output grid will contain n non-zero grid cells, randomly distributed throughout the raster grid, and a background value of zero. This tool is useful when performing statistical analyses on raster images when you wish to obtain a random sample of data.

Only valid, non-nodata, cells in the base raster will be sampled.

Function Signature

def random_sample(self, base_raster: Raster = None, num_samples: int = 1000) -> Raster: ...

range_filter

This tool performs a range filter on an input image (input). A range filter assigns to each cell in the output grid the range (maximum - minimum) of the values contained within a moving window centred on each grid cell.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

See Also

total_filter

Function Signature

def range_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

raster_area

This tools estimates the area of each category, polygon, or patch in an input raster. The input raster must be categorical in data scale. Rasters with floating-point cell values are not good candidates for an area analysis. The user must specify whether the output is given in grid cells or map units (units). Map Units are physical units, e.g. if the rasters's scale is in metres, areas will report in square-metres. Notice that square-metres can be converted into hectares by dividing by 10,000 and into square-kilometres by dividing by 1,000,000. If the input raster is in geographic coordinates (i.e. latitude and longitude) a warning will be issued and areas will be estimated based on per-row calculated degree lengths.

The tool can be run with a raster output (output), a text output (out_text), or both. If niether outputs are specified, the tool will automatically output a raster named area.tif.

Zero values in the input raster may be excluded from the area analysis if the zero_back flag is used.

To calculate the area of vector polygons, use the polygon_area tool instead.

See Also

polygon_area, raster_histogram

Function Signature

def raster_area(self, raster: Raster, units: str = "map units", zero_background: bool = False) -> Tuple[Raster, str]: ...

raster_calculator

The raster_calculator tool can be used to perform a complex mathematical operations on one or more input raster images on a cell-to-cell basis. The user inputs an expression and a list of input rasters (input_rasters), specified in the same order as the rasters contained within the statement. Rasters are treated like variables (that change value with each grid cell) and are specified within the statement as arbitrarily named variables contained within either double or single quotation marks (e.g. "DEM" > 500.0). The order of raster variables must match the order of rasters within the input_rasters list.**Note, all input rasters must share the same number of rows and columns and spatial extent. Use the resample tool if this is not the case to convert the one raster's grid resolution to the others.

Example

(band3, band4) = wbe.read_rasters('band3.tif', 'band4.tif')
result = wbe.raster_calculator("('nir' - 'red') / ('nir' + 'red')", [band4, band3])
wbe.write_raster(result, 'result.tif', True)

The mathematical expression supports all of the standard algebraic unary and binary operators (+ - * / ^ %), as well as comparisons (< <= == != >= >) and logical operators (&& ||) with short-circuit support. The order of operations, from highest to lowest is as follows.

Listed in order of precedence:

OrderSymbolDescription
(Highest Precedence)^Exponentiation
%Modulo
/Division
*Multiplication
-Subtraction
+Addition
== != < <= >= >Comparisons (all have equal precedence)
&& andLogical AND with short-circuit
(Lowest Precedence)|| orLogical OR with short-circuit

Several common mathematical functions are also available for use in the input statement. For example:

 * log(base=10, val) -- Logarithm with optional 'base' as first argument.
 If not provided, 'base' defaults to '10'.
 Example: log(100) + log(e(), 100)

 * e()  -- Euler's number (2.718281828459045)
 * pi() -- π (3.141592653589793)

 * int(val)
 * ceil(val)
 * floor(val)
 * round(modulus=1, val) -- Round with optional 'modulus' as first argument.
     Example: round(1.23456) == 1 && round(0.001, 1.23456) == 1.235

 * abs(val)
 * sign(val)

 * min(val, ...) -- Example: min(1, -2, 3, -4) == -4
 * max(val, ...) -- Example: max(1, -2, 3, -4) == 3

 * sin(radians)    * asin(val)
 * cos(radians)    * acos(val)
 * tan(radians)    * atan(val)
 * sinh(val)       * asinh(val)
 * cosh(val)       * acosh(val)
 * tanh(val)       * atanh(val)

Notice that the constants pi and e must be specified as functions, pi() and e(). A number of global variables are also available to build conditional statements. These include the following:

Special Variable Names For Use In Conditional Statements:

NameDescription
nodataAn input raster's NoData value.
nullSame as nodata.
minvalueAn input raster's minimum value.
maxvalueAn input raster's maximum value.
rowsThe input raster's number of rows.
columnsThe input raster's number of columns.
rowThe grid cell's row number.
columnThe grid cell's column number.
rowyThe row's y-coordinate.
columnxThe column's x-coordinate.
northThe input raster's northern coordinate.
southThe input raster's southern coordinate.
eastThe input raster's eastern coordinate.
westThe input raster's western coordinate.
cellsizexThe input raster's grid resolution in the x-direction.
cellsizeyThe input raster's grid resolution in the y-direction.
cellsizeThe input raster's average grid resolution.

The special variable names are case-sensitive. If there are more than one raster inputs used in the statement, the functional forms of the nodata, null, minvalue, and maxvalue variables should be used, e.g. nodata("InputRaster"), otherwise the value is assumed to specify the attribute of the first raster in the statement. The following are examples of valid statements:

 "raster" != 300.0

 "raster" >= (minvalue + 35.0)

 ("raster1" >= 25.0) && ("raster2" <= 75.0) -- Evaluates to 1 where both conditions are true.

 tan("raster" * pi() / 180.0) > 1.0

 "raster" == nodata

Any grid cell in the input rasters containing the NoData value will be assigned NoData in the output raster, unless a NoData grid cell value allows the statement to evaluate to True (i.e. the mathematical expression includes the nodata value).

See Also

ConditionalEvaluation

Function Signature

def raster_calculator(self, expression: str, input_rasters: List[Raster]) -> Raster: ...

raster_cell_assignment

This tool can be used to create a new raster with the same coordinates and dimensions (i.e. rows and columns) as an existing base image. Grid cells in the new raster will be assigned either the row or column number or the x- or y-coordinate, depending on the selected option (assign flag). The user must also specify the name of the base image (input).

See Also

NewRasterFromBase

Function Signature

def raster_cell_assignment(self, raster: Raster, what_to_assign: str = "column") -> Raster: ...

raster_histogram

This tool produces a histogram (i.e. a frequency distribution graph) for the values contained within an input raster file (input). The histogram will be embedded within an output (output) HTML file, which should be automatically displayed after the tool has completed.

See Also

attribute_histogram

Function Signature

def raster_histogram(self, raster: Raster, output_html_file: str) -> None: ...

raster_perimeter

This tool can be used to measure the length of the perimeter of polygon features in a raster layer. The user must specify the name of the input raster file (input) and optionally an output raster (output), which is the raster layer containing the input features assigned the perimeter length. The user may also optionally choose to output text data (out_text). Raster-based perimeter estimation uses the accurate, anti-aliasing algorithm of Prashker (2009).

The input file must be of a categorical data type, containing discrete polygon features that have been assigned unique identifiers. Such rasters are often created by region-grouping (clump) a classified raster.

Reference

Prashker, S. (2009) An anti-aliasing algorithm for calculating the perimeter of raster polygons. Geotec, Ottawa and Geomtics Atlantic, Wolfville, NS.

See Also

raster_area, clump

Function Signature

def raster_perimeter(self, raster: Raster, units: str = "map units", zero_background: bool = False) -> Tuple[Raster, str]: ...

raster_streams_to_vector

This tool converts a raster stream file into a vector file. The user must specify an input raster streams file (streams), and an input D8 flow pointer file (d8_pointer). Streams in the input raster streams file are denoted by cells containing any positive, non-zero integer. A field in the output vector's database file, called STRM_VAL, will correspond to this positive integer value. The database file will also have a field for the length of each link in the stream network. The flow pointer file must be calculated from a DEM with all topographic depressions and flat areas removed and must be calculated using the D8 flow pointer algorithm (d8_pointer). The output vector will contain PolyLine features.

See Also

rasterize_streams, raster_to_vector_lines

Function Signature

def raster_streams_to_vector(self, streams: Raster, d8_pointer: Raster, esri_pointer: bool = False) -> Vector: ...

raster_summary_stats

This tool outputs distribution summary statistics for input raster images (input). The distribution statistics include the raster minimum, maximum, range, total, mean, variance, and standard deviation. These summary statistics are output to the system stdout.

The following is an example of the summary report:

*********************************
* Welcome to RasterSummaryStats *
*********************************
Reading data...

Number of non-nodata grid cells: 32083559
Number of nodata grid cells: 3916441
Image minimum: 390.266357421875
Image maximum: 426.0322570800781
Image range: 35.765899658203125
Image total: 13030334843.332886
Image average: 406.13745012929786
Image variance: 31.370027239143383
Image standard deviation: 5.600895217654351

See Also

raster_histogram, zonal_statistics

Function Signature

def raster_summary_stats(self, input: Raster) -> str: ...

raster_to_vector_lines

This tool converts raster lines features into a vector of the POLYLINE VectorGeometryType. Grid cells associated with line features will contain non-zero, non-NoData cell values. The algorithm requires three passes of the raster. The first pass counts
the number of line neighbours of each line cell; the second pass traces line segments starting from line ends (i.e. line cells with only one neighbouring line cell); lastly, the final pass traces any remaining line segments, which are likely forming closed loops (and therefore do not have line ends).

If the line raster contains streams, it is preferable to use the raster_streams_to_vector instead. This tool will use knowledge of flow directions to ensure connections between stream segments at confluence sites, whereas raster_to_vector_lines will not.

See Also

raster_to_vector_polygons, raster_to_vector_points, raster_streams_to_vector

Function Signature

def raster_to_vector_lines(self, raster: Raster) -> Vector: ...

raster_to_vector_points

Converts a raster data set to a vector of the POINT VectorGeometryType. The user must specify the name of a raster file (input) and the name of the output vector (output). Points will correspond with grid cell centre points. All grid cells containing non-zero, non-NoData values will be considered a point. The vector's attribute table will contain a field called 'VALUE' that will contain the cell value for each point feature.

See Also

raster_to_vector_polygons, raster_to_vector_lines

Function Signature

def raster_to_vector_points(self, raster: Raster) -> Vector: ...

raster_to_vector_polygons

Converts a raster data set to a vector of the POLYGON geometry type. The user must specify the name of a raster file (input) and the name of the output (output) vector. All grid cells containing non-zero, non-NoData values will be considered part of a polygon feature. The vector's attribute table will contain a field called 'VALUE' that will contain the cell value for each polygon feature, in addition to the standard feature ID (FID) attribute.

See Also

raster_to_vector_points, raster_to_vector_lines

Function Signature

def raster_to_vector_polygons(self, raster: Raster) -> Vector: ...

rasterize_streams

This tool can be used rasterize an input vector stream network (streams) using on Lindsay (2016) method. The user inputs an existing raster (base_raster), from which the output raster's grid resolution is determined.

Reference

Lindsay JB. 2016. The practice of DEM stream burning revisited. Earth Surface Processes and Landforms, 41(5): 658–668. DOI: 10.1002/esp.3888

See Also

raster_streams_to_vector

Function Signature

def rasterize_streams(self, streams: Vector, base_raster: Raster = None, zero_background: bool = False, use_feature_id: bool = False) -> Raster: ...

read_lidar

Returns a new Lidar object, read from a path-file string.

Parameters

  • file_name: str - The file name. If file_name does not contain the full file path, the file will be read from the Whitebox working directory.

Example

import whitebox_workflows

wbe = whitebox_workflows.WbEnvironment()
my_lidar = wbe.read_lidar("path/containing/file/file_name.laz")

read_lidars

Reads multiple LiDAR files into memory at once, returning a list of Lidar objects.

Parameters

  • file_names: List[str] - The file names. If a file name does not contain the full file path, the file will be read from the Whitebox working directory.

Example

import whitebox_workflows

wbe = whitebox_workflows.WbEnvironment()
wbe.working_directory = '/path/to/data'

# Notice that you can use tuple destructuring on the resulting list of rasters
tile1, tile2, tile3 = wbe.read_lidars(['tile1.laz', 'tile2.laz', 'tile3.laz'])

read_raster

Returns a new Raster object, read into memory from a path-file string.

Parameters

  • file_name: str - The file name. If file_name does not contain the full file path, the file will be read from the Whitebox working directory.

Example

import whitebox_workflows

wbe = whitebox_workflows.WhiteboxEnvironment()
my_raster = we.read_raster("path/containing/file/file_name.tif")

read_rasters

Reads multiple raster files into memory at once, returning a list of Raster objects.

Parameters

  • file_names: List[str] - The list of file name strings. If any of the files do not contain the full file path, the file will be read from the Whitebox working directory.

Example

import whitebox_workflows

wbe = whitebox_workflows.WhiteboxEnvironment()
wbe.working_directory = '/path/to/data'

# Notice that you can use tuple destructuring on the resulting list of rasters
band1, band2, band3 = wbe.read_rasters(['band1.tif', 'band2.tif', 'band3.tif'])

read_vector

Reads a vector from disc into an in-memory Vector object.

Parameters

  • file_name: str - The file name. If file_name does not contain the full file path, the file will be read from the Whitebox working directory.

read_vectors

Reads multiple vectors from file into a list of in-memory Vector objects.

Parameters

  • file_names: List[str] - The list of file names. If any of the file names do not contain the full file path, the file will be read from the Whitebox working directory.

reciprocal

This tool creates a new raster (output) in which each grid cell is equal to one divided by the grid cell values in the input raster image (input). NoData values in the input image will be assigned NoData values in the output image.

Function Signature

def reciprocal(self, raster: Raster) -> Raster: ...

reclass

This tool creates a new raster in which the value of each grid cell is determined by an input raster (input) and a collection of user-defined classes. The user must specify the New value, the From value, and the To Just Less Than value of each class triplet of the reclass string. Classes must be mutually exclusive, i.e. non-overlapping. For example:

Function Signature

def reclass(self, raster: Raster, reclass_values: List[List[float]], assign_mode: bool = False) -> Raster: ...

reclass_equal_interval

This tool reclassifies the values in an input raster (input) file based on an equal-interval scheme, where the user must specify the reclass interval value (interval), the starting value (start_val), and optionally, the ending value (end_val). Grid cells containing values that fall outside of the range defined by the starting and ending values, will be assigned their original values in the output grid. If the user does not specify an ending value, the tool will assign a very large positive value.

See Also

reclass

Function Signature

def reclass_equal_interval(self, raster: Raster, interval_size: float, start_value: float = float('-inf'), end_value: float = float('inf')) -> Raster: ...

rectangular_grid_from_raster_base

This tool can be used to create a rectangular vector grid. The extent of the rectangular grid is based on the extent of an input base raster (base). The user may also specify the origin of the grid (xorig and yorig, defaults are 0.0) and the grid cell width and height (width and height).

See Also

rectangular_grid_from_vector_base, hexagonal_grid_from_raster

Function Signature

def rectangular_grid_from_raster_base(self, base: Raster, width: float, height: float, x_origin: float = 0.0, y_origin: float = 0.0) -> Vector: ...

rectangular_grid_from_vector_base

This tool can be used to create a rectangular vector grid. The extent of the rectangular grid is based on the extent of an input base vector (base). The user may also specify the origin of the grid (xorig and yorig, defaults are 0.0) and the grid cell width and height (width and height).

See Also

rectangular_grid_from_raster_base, hexagonal_grid_from_vector

Function Signature

def rectangular_grid_from_vector_base(self, base: Vector, width: float, height: float, x_origin: float = 0.0, y_origin: float = 0.0) -> Vector: ...

reinitialize_attribute_table

Reinitializes a vector's attribute table deleting all fields but the feature ID (FID). Caution: this tool overwrites the input file's attribute table.

Function Signature

def reinitialize_attribute_table(self, input: Vector) -> None: ...

related_circumscribing_circle

This tool can be used to calculate the related circumscribing circle (Mcgarigal et al. 2002) for vector polygon features. The related circumscribing circle values calculated for each vector polygon feature will be placed in the accompanying attribute table as a new field (RC_CIRCLE).

Related circumscribing circle (RCC) is defined as:

RCC = 1 - A / Ac

Where A is the polygon's area and Ac the area of the smallest circumscribing circle.

Theoretically, related_circumscribing_circle ranges from 0 to 1, where a value of 0 indicates a circular polygon and a value of 1 indicates a highly elongated shape. The circumscribing circle provides a measure of polygon elongation. Unlike the elongation_ratio, however, it does not provide a measure of polygon direction in addition to overall elongation. Like the elongation_ratio and linearity_index, related_circumscribing_circle is not an adequate measure of overall polygon narrowness, because a highly sinuous but narrow patch will have a low related circumscribing circle index owing to the compact nature of these polygon.

Note: Holes are excluded from the area calculation of polygons.

Function Signature

def related_circumscribing_circle(self, input: Vector) -> Vector: ...

relative_aspect

This tool creates a new raster in which each grid cell is assigned the terrain aspect relative to a user-specified direction (azimuth). Relative terrain aspect is the angular distance (measured in degrees) between the land-surface aspect and the assumed regional wind azimuth (Bohner and Antonic, 2007). It is bound between 0-degrees (windward direction) and 180-degrees (leeward direction). Relative terrain aspect is the simplest of the measures of topographic exposure to wind, taking into account terrain orientation only and neglecting the influences of topographic shadowing by distant landforms and the deflection of wind by topography.

The user must input a digital elevation model (DEM) (dem) and an azimuth (i.e. a wind direction). The Z Conversion Factor (zfactor) is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z Conversion Factor.

Reference

Böhner, J., and Antonić, O. (2009). Land-surface parameters specific to topo-climatology. Developments in Soil Science, 33, 195-226.

See Also

aspect

Function Signature

def relative_aspect(self, dem: Raster, azimuth: float = 0.0, z_factor: float = 1.0) -> Raster: ...

relative_stream_power_index

This tool can be used to calculate the relative stream power (RSP) index. This index is directly related to the stream power if the assumption can be made that discharge is directly proportional to upslope contributing area (As; sca). The index is calculated as:

RSP = Asp × tan(β)

where As is the specific catchment area (i.e. the upslope contributing area per unit contour length) estimated using one of the available flow accumulation algorithms; β is the local slope gradient in degrees (slope); and, p (exponent) is a user-defined exponent term that controls the location-specific relation between contributing area and discharge. Notice that As must not be log-transformed prior to being used; As is commonly log-transformed to enhance visualization of the data. The slope raster can be created from the base digital elevation model (DEM) using the slope tool. The input images must have the same grid dimensions.

Reference

Moore, I. D., Grayson, R. B., and Ladson, A. R. (1991). Digital terrain modelling: a review of hydrological, geomorphological, and biological applications. Hydrological processes, 5(1), 3-30.

See Also

sediment_transport_index, slope, D8FlowAccumulation DInfFlowAccumulation, FD8FlowAccumulation

Function Signature

def relative_stream_power_index(self, specific_catchment_area: Raster, slope: Raster, exponent: float = 1.0) -> Raster: ...

relative_topographic_position

Relative topographic position (RTP) is an index of local topographic position (i.e. how elevated or low-lying a site is relative to its surroundings) and is a modification of percent elevation range (PER; percent_elev_range) and accounts for the elevation distribution. Rather than positioning the central cell's elevation solely between the filter extrema, RTP is a piece-wise function that positions the central elevation relative to the minimum (zmin), mean (μ), and maximum values (zmax), within a local neighbourhood of a user-specified size (filterx, filtery), such that:

RTP = (z0 − μ) / (μ − zmin), if z0 < μ

OR

RTP = (z0 − μ) / (zmax - μ), if z0 >= μ 

The resulting index is bound by the interval [−1, 1], where the sign indicates if the cell is above or below than the filter mean. Although RTP uses the mean to define two linear functions, the reliance on the filter extrema is expected to result in sensitivity to outliers. Furthermore, the use of the mean implies assumptions of unimodal and symmetrical elevation distribution.

In many cases, Elevation Percentile (ElevPercentile) and deviation from mean elevation (DevFromMeanElev) provide more suitable and robust measures of relative topographic position.

Reference

Newman, D. R., Lindsay, J. B., and Cockburn, J. M. H. (2018). Evaluating metrics of local topographic position for multiscale geomorphometric analysis. Geomorphology, 312, 40-50.

See Also

DevFromMeanElev, DiffFromMeanElev, ElevPercentile, percent_elev_range

Function Signature

def relative_topographic_position(self, dem: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

remove_duplicates

This tool removes duplicate points from a LiDAR data set. Duplicates are determined by their x, y, and optionally (include_z) z coordinates.

See Also

eliminate_coincident_points

Function Signature

def remove_duplicates(self, input: Lidar, include_z: bool = False) -> Lidar: ...

remove_off_terrain_objects

This tool can be used to create a bare-earth DEM from a fine-resolution digital surface model. The tool is typically applied to LiDAR DEMs which frequently contain numerous off-terrain objects (OTOs) such as buildings, trees and other vegetation, cars, fences and other anthropogenic objects. The algorithm works by finding and removing steep-sided peaks within the DEM. All peaks within a sub-grid, with a dimension of the user-specified maximum OTO size (filter), in pixels, are identified and removed. Each of the edge cells of the peaks are then examined to see if they have a slope that is less than the user-specified minimum OTO edge slope (slope) and a back-filling procedure is used. This ensures that OTOs are distinguished from natural topographic features such as hills. The DEM is preprocessed using a white top-hat transform, such that elevations are normalized for the underlying ground surface.

Note that this tool is appropriate to apply to rasterized LiDAR DEMs. Use the lidar_ground_point_filter tool to remove or classify OTOs within a LiDAR point-cloud.

Reference

J.B. Lindsay (2018) A new method for the removal of off-terrain objects from LiDAR-derived raster surface models. Available online, DOI: 10.13140/RG.2.2.21226.62401

See Also

map_off_terrain_objects, tophat_transform, lidar_ground_point_filter

Function Signature

def remove_off_terrain_objects(self, dem: Raster, filter_size: int = 11, slope_threshold: float = 15.0) -> Raster: ...

remove_polygon_holes

This tool can be used to remove holes from the features within a vector polygon file. The user must specify the name of the input vector file, which must be of a polygon VectorGeometryType, and the name of the output file.

Function Signature

def remove_polygon_holes(self, input: Vector) -> Vector: ...

remove_short_streams

This tool can be used to remove stream links in a stream network that are shorter than a user-specified length (min_length). The user must input a streams raster image (streams_raster) and D8 pointer (flow direction) image (d8_pntr). Stream cells are designated in the streams raster as all positive, nonzero values. Thus all non-stream or background grid cells are commonly assigned either zeros or NoData values. The pointer raster is used to traverse the stream network and should only be created using the D8 algorithm (d8_pointer).

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, the user must specify esri_pntr=True.

See Also

extract_streams, d8_pointer

Function Signature

def remove_short_streams(self, d8_pntr: Raster, streams_raster: Raster, min_length: float = 0.0, esri_pntr: bool = False) -> Raster: ...

remove_spurs

This image processing tool removes small irregularities (i.e. spurs) on the boundaries of objects in a Boolean input raster image (input). This operation is sometimes called pruning. Remove Spurs is a useful tool for cleaning an image before performing a line thinning operation. In fact, the input image need not be truly Boolean (i.e. contain only 1's and 0's). All non-zero, positive values are considered to be foreground pixels while all zero valued cells are considered background pixels.

Note: Unlike other filter-based operations in WhiteboxTools, this algorithm can't easily be parallelized because the output raster must be read and written to during the same loop.

See Also

line_thinning

Function Signature

def remove_spurs(self, raster: Raster, max_iterations: int = 10) -> Raster: ...

resample

This tool can be used to modify the grid resolution of one or more rasters. The user specifies the names of one or more input rasters (inputs). The resolution of the output raster is determined either using a specified cell_size parameter, in which case the output extent is determined by the combined extent of the inputs, or by an optional base raster (base), in which case the output raster spatial extent matches that of the base file. This operation is similar to the mosaic tool, except that resample modifies the output resolution. The resample tool may also be used with a single input raster (when the user wants to modify its spatial resolution, whereas, mosaic always includes multiple inputs.

If the input source images are more extensive than the base image (if optionally specified), these areas will not be represented in the output image. Grid cells in the output image that are not overlapping with any of the input source images will not be assigned the NoData value, which will be the same as the first input image. Grid cells in the output image that overlap with multiple input raster cells will be assigned the last input value in the stack. Thus, the order of input images is important.

See Also

mosaic

Function Signature

def resample(self, input_rasters: List[Raster], cell_size: float = 0.0, base_raster: Raster = None, method: str = "cc") -> Raster: ...

rescale_value_range

Function Signature

def rescale_value_range(self, raster: Raster, out_min_val: float, out_max_val: float, clip_min: float = float('inf'), clip_max: float = float('-inf')) -> Raster: ...

rgb_to_ihs

This tool transforms three raster images of multispectral data (red, green, and blue channels) into their equivalent intensity, hue, and saturation (IHS; sometimes HSI or HIS) images. Intensity refers to the brightness of a color, hue is related to the dominant wavelength of light and is perceived as color, and saturation is the purity of the color (Koutsias et al., 2000). There are numerous algorithms for performing a red-green-blue (RGB) to IHS transformation. This tool uses the transformation described by Haydn (1982). Note that, based on this transformation, the output IHS values follow the ranges:

0 < I < 1

0 < H < 2PI

0 < S < 1

The user must specify the names of the red, green, and blue images (red, green, blue). Importantly, these images need not necessarily correspond with the specific regions of the electromagnetic spectrum that are red, green, and blue. Rather, the input images are three multispectral images that could be used to create a RGB color composite. The user must also specify the names of the output intensity, hue, and saturation images (intensity, hue, saturation). Image enhancements, such as contrast stretching, are often performed on the IHS components, which are then inverse transformed back in RGB components to then create an improved color composite image.

References

Haydn, R., Dalke, G.W. and Henkel, J. (1982) Application of the IHS color transform to the processing of multisensor data and image enhancement. Proc. of the Inter- national Symposium on Remote Sensing of Arid and Semiarid Lands, Cairo, 599-616.

Koutsias, N., Karteris, M., and Chuvico, E. (2000). The use of intensity-hue-saturation transformation of Landsat-5 Thematic Mapper data for burned land mapping. Photogrammetric Engineering and Remote Sensing, 66(7), 829-840.

See Also

ihs_to_rgb, balance_contrast_enhancement, direct_decorrelation_stretch

Function Signature

def rgb_to_ihs(self, red: Raster, green: Raster, blue: Raster, composite: Raster) -> Tuple[Raster, Raster, Raster]: ...

rho8_flow_accum

This tool is used to generate a flow accumulation grid (i.e. contributing area) using the Fairfield and Leymarie (1991) flow algorithm, often called Rho8. Like the D8 flow method, this algorithm is an examples of a single-flow-direction (SFD) method because the flow entering each grid cell is routed to only one downslope neighbour, i.e. flow divergence is not permitted. The user must specify the name of the input file (input), which may be either a digital elevation model (DEM) or a Rho8 pointer file (see rho8_pointer). If a DEM is input, it must have been hydrologically corrected to remove all spurious depressions and flat areas. DEM pre-processing is usually achieved using either the breach_depressions_least_cost (also breach_depressions_least_cost) or fill_depressions tool.

In addition to the input and output (output)files, the user must also specify the output type (out_type). The output flow-accumulation can be: 1) cells (i.e. the number of inflowing grid cells), catchment area (i.e. the upslope area), or specific contributing area (i.e. the catchment area divided by the flow width). The default value is specific contributing area. The user must also specify whether the output flow-accumulation grid should be log-tranformed (log), i.e. the output, if this option is selected, will be the natural-logarithm of the accumulated flow value. This is a transformation that is often performed to better visualize the contributing area distribution. Because contributing areas tend to be very high along valley bottoms and relatively low on hillslopes, when a flow-accumulation image is displayed, the distribution of values on hillslopes tends to be 'washed out' because the palette is stretched out to represent the highest values. Log-transformation provides a means of compensating for this phenomenon. Importantly, however, log-transformed flow-accumulation grids must not be used to estimate other secondary terrain indices, such as the wetness index (wetness_index), or relative stream power index (StreamPowerIndex).

If a Rho8 pointer is used as the input raster, the user must specify this (pntr). Similarly, if a pointer input is used and the pointer follows the Esri pointer convention, rather than the default WhiteboxTools convension for pointer files, then this must also be specified (esri_pntr).

Reference

Fairfield, J., and Leymarie, P. 1991. Drainage networks from grid digital elevation models. Water Resources Research, 27(5), 709-717.

See Also

rho8_pointer, D8FlowAccumulation, qin_flow_accumulation, FD8FlowAccumulation, DInfFlowAccumulation, MDInfFlowAccumulation, wetness_index

Function Signature

def rho8_flow_accum(self, raster: Raster, out_type: str = "sca", log_transform: bool = False, clip: bool = False, input_is_pointer: bool = False, esri_pntr: bool = False) -> Raster: ...

rho8_pointer

This tool is used to generate a flow pointer grid (i.e. flow direction) using the stochastic Rho8 (J. Fairfield and P. Leymarie, 1991) algorithm. Like the D8 flow algorithm (d8_pointer), Rho8 is a single-flow-direction (SFD) method because the flow entering each grid cell is routed to only one downslope neighbour, i.e. flow divergence is not permitted. The user must specify the name of a digital elevation model (DEM) file (dem) that has been hydrologically corrected to remove all spurious depressions and flat areas (breach_depressions_least_cost, fill_depressions). The output of this tool (output) is often used as the input to the Rho8FlowAccumulation tool.

By default, the Rho8 flow pointers use the following clockwise, base-2 numeric index convention:

...
641281
3202
1684

Notice that grid cells that have no lower neighbours are assigned a flow direction of zero. In a DEM that has been pre-processed to remove all depressions and flat areas, this condition will only occur along the edges of the grid. If the pointer file contains ESRI flow direction values instead, the esri_pntr parameter must be specified.

Grid cells possessing the NoData value in the input DEM are assigned the NoData value in the output image.

Memory Usage

The peak memory usage of this tool is approximately 10 bytes per grid cell.

References

Fairfield, J., and Leymarie, P. 1991. Drainage networks from grid digital elevation models. Water Resources Research, 27(5), 709-717.

See Also

Rho8FlowAccumulation, d8_pointer, fd8_pointer, DInfPointer, breach_depressions_least_cost, fill_depressions

Function Signature

def rho8_pointer(self, dem: Raster, esri_pntr: bool = False) -> Raster: ...

roberts_cross_filter

This tool performs Robert's Cross edge-detection filter on a raster image. The roberts_cross_filter
is similar to the sobel_filter and prewitt_filter, in that it identifies areas of high slope in the input image through the calculation of slopes in the x and y directions. A Robert's Cross filter uses the following 2 × 2 schemes to calculate slope magnitude, |G|:

..
P1P2
P3P4

|G| = |P1 - P4| + |P2- P3|

Note, the filter is centered on pixel P1 and P2, P3, and P4 are the neighbouring pixels towards the east, south, and south-east respectively.

The output image may be overwhelmed by a relatively small number of high-valued pixels, stretching the palette. The user may therefore optionally clip the output image distribution tails by a specified amount (clip) for improved visualization.

Reference

Fisher, R. 2004. Hypertext Image Processing Resources 2 (HIPR2). Available online: http://homepages.inf.ed.ac.uk/rbf/HIPR2/roberts.htm

See Also

sobel_filter, prewitt_filter

Function Signature

def roberts_cross_filter(self, raster: Raster, clip_amount: float = 0.0) -> Raster: ...

root_mean_square_error

This tool calculates the root-mean-square-error (RMSE) or root-mean-square-difference (RMSD) from two input rasters. If the two input rasters possess the same number of rows and columns, the RMSE is calucated on a cell-by-cell basis, otherwise bilinear resampling is used. In addition to RMSE, the tool also reports other common accuracy statistics including the mean verical error, the 95% confidence limit (RMSE x 1.96), and the 90% linear error (LE90), which is the 90% percentile of the residuals between two raster surfaces. The LE90 is the most robust of the reported accuracy statistics when the residuals are non-Gaussian. The LE90 requires sorting the residual values, which can be a relatively slow operation for larger rasters.

See Also

paired_sample_t_test, wilcoxon_signed_rank_test

Function Signature

def root_mean_square_error(self, input: Raster, reference: Raster) -> str: ...

ruggedness_index

The terrain ruggedness index (TRI) is a measure of local topographic relief. The TRI calculates the root-mean-square-deviation (RMSD) for each grid cell in a digital elevation model (DEM), calculating the residuals (i.e. elevation differences) between a grid cell and its eight neighbours. Notice that, unlike the output of this tool, the original Riley et al. (1999) TRI did not normalize for the number of cells in the local window (i.e. it is a root-square-deviation only). However, using the mean has the advantage of allowing for the varying number of neighbouring cells along the grid edges and in areas bordering NoData cells. This modification does however imply that the output of this tool cannot be directly compared with the index ranges of level to extremely rugged terrain provided in Riley et al. (1999)

Reference

Riley, S. J., DeGloria, S. D., and Elliot, R. (1999). Index that quantifies topographic heterogeneity. Intermountain Journal of Sciences, 5(1-4), 23-27.

See Also

relative_topographic_position, DevFromMeanElev

Function Signature

def ruggedness_index(self, input: Raster) -> Raster: ...

scharr_filter

This tool performs a Scharr edge-detection filter on a raster image. The Scharr filter is similar to the sobel_filter and prewitt_filter, in that it identifies areas of high slope in the input image through the calculation of slopes in the x and y directions. A 3 × 3 Scharr filter uses the following schemes to calculate x and y slopes:

X-direction slope

...
30-3
100-10
30-3

Y-direction slope

...
3103
000
-3-10-3

Each grid cell in the output image is assigned the square-root of the squared sum of the x and y slopes.

The output image may be overwhelmed by a relatively small number of high-valued pixels, stretching the palette. The user may therefore optionally clip the output image distribution tails by a specified amount (clip) for improved visualization.

See Also

sobel_filter, prewitt_filter

Function Signature

def scharr_filter(self, raster: Raster, clip_tails: float = 0.0) -> Raster: ...

sediment_transport_index

This tool calculates the sediment transport index, or sometimes, length-slope (LS) factor, based on input specific contributing area (As, i.e. the upslope contributing area per unit contour length; sca) and slope gradient (β, measured in degrees; slope) rasters. Moore et al. (1991) state that the physical potential for sheet and rill erosion in upland catchments can be evaluated by the product R K LS, a component of the Universal Soil Loss Equation (USLE), where R is a rainfall and runoff erosivity factor, K is a soil erodibility factor, and LS is the length-slope factor that accounts for the effects of topography on erosion. To predict erosion at a point in the landscape the LS factor can be written as:

LS = (n + 1)(As / 22.13)n(sin(β) / 0.0896)m

where n = 0.4 (sca_exponent) and m = 1.3 (slope_exponent) in its original formulation.

This index is derived from unit stream-power theory and is sometimes used in place of the length-slope factor in the revised universal soil loss equation (RUSLE) for slope lengths less than 100 m and slope less than 14 degrees. Like many hydrological land-surface parameters sediment_transport_index assumes that contributing area is directly related to discharge. Notice that As must not be log-transformed prior to being used; As is commonly log-transformed to enhance visualization of the data. Also, As can be derived using any of the available flow accumulation tools, alghough better results usually result from application of multiple-flow direction algorithms such as DInfFlowAccumulation and FD8FlowAccumulation. The slope raster can be created from the base digital elevation model (DEM) using the slope tool. The input images must have the same grid dimensions.

Reference

Moore, I. D., Grayson, R. B., and Ladson, A. R. (1991). Digital terrain modelling: a review of hydrological, geomorphological, and biological applications. Hydrological processes, 5(1), 3-30.

See Also

StreamPowerIndex, DInfFlowAccumulation, FD8FlowAccumulation

Function Signature

def sediment_transport_index(self, specific_catchment_area: Raster, slope: Raster, sca_exponent: float = 0.4, slope_exponent: float = 1.3) -> Raster: ...

select_tiles_by_polygon

This tool copies LiDAR tiles overlapping with a polygon into an output directory. In actuality, the tool performs point-in-polygon operations, using the four corner points, the center point, and the four mid-edge points of each LiDAR tile bounding box and the polygons. This representation of overlapping geometry aids with performance. This approach generally works well when the polygon size is large relative to the LiDAR tiles. If, however, the input polygon is small relative to the tile size, this approach may miss some copying some tiles. It is advisable to buffer the polygon if this occurs.

See Also

lidar_tile_footprint

Function Signature

def select_tiles_by_polygon(self, input_directory: str, output_directory: str, polygons: Vector) -> None: ...

set_nodata_value

This tool will re-assign a user-defined background value in an input raster image the NoData value. More precisely, the NoData value will be changed to the specified background value and any existing grid cells containing the previous NoData value, if it had been defined, will be changed to this new value. Most WhiteboxTools tools recognize NoData grid cells and treat them specially. NoData grid cells are also often displayed transparently by GIS software. The user must specify the names of the input and output rasters and the background value. The default background value is zero, although any numeric value is possible.

This tool differs from the ModifyNoDataValue tool in that it simply updates the NoData value in the raster header, without modifying pixel values. The ModifyNoDataValue tool will update the value in the header, and then modify each existing NoData pixel to contain this new value. Also, set_nodata_value does not overwrite the input file, while the ModifyNoDataValue tool does.

See Also

ModifyNoDataValue, convert_nodata_to_zero, IsNoData

Function Signature

def set_nodata_value(self, raster: Raster, back_value: float = 0.0) -> Raster: ...

shape_complexity_index_raster

This tools calculates a type of shape complexity index for raster objects. The index is equal to the average number of intersections of the group of vertical and horizontal transects passing through an object. Simple objects will have a shape complexity index of 1.0 and more complex shapes, including those containing numerous holes or are winding in shape, will have higher index values. Objects in the input raster (input) are designated by their unique identifiers. Identifier values should be positive, non-zero whole numbers.

See Also

ShapeComplexityIndex, boundary_shape_complexity

Function Signature

def shape_complexity_index_raster(self, raster: Raster) -> Raster: ...

shape_complexity_index_vector

This tool provides a measure of overall polygon shape complexity, or irregularity, for vector polygons. Several shape indices have been created to compare a polygon's shape to simple Euclidean shapes (e.g. circles, squares, etc.). One of the problems with this approach is that it inherently convolves the characteristics of polygon complexity and elongation. The Shape Complexity Index (SCI) was developed as a parameter for assessing the complexity of a polygon that is independent of its elongation.

SCI relates a polygon's shape to that of an encompassing convex hull. It is defined as:

SCI = 1 - A / Ah

Where A is the polygon's area and Ah is the area of the convex hull containing the polygon. Convex polygons, i.e. those that do not contain concavities or holes, have a value of 0. As the shape of the polygon becomes more complex, the SCI approaches 1. Note that polygon shape complexity also increases with the greater number of holes (i.e. islands), since holes have the effect of reducing the lake area.

The SCI values calculated for each vector polygon feature will be placed in the accompanying database file (.dbf) as a complexity field (COMPLEXITY).

See Also

shape_complexity_index_raster

Function Signature

def shape_complexity_index_vector(self, input: Vector) -> Vector: ...

shreve_stream_magnitude

This tool can be used to assign the Shreve stream magnitude to each link in a stream network. Stream ordering is often used in hydro-geomorphic and ecological studies to quantify the relative size and importance of a stream segment to the overall river system. There are several competing stream ordering schemes. Shreve stream magnitude is equal to the number of headwater links upstream of each link. Headwater stream links are assigned a magnitude of one.

The user must input a streams raster image (streams_raster) and D8 pointer (flow direction) image (d8_pntr). Stream cells are designated in the streams raster as all positive, nonzero values. Thus all non-stream or background grid cells are commonly assigned either zeros or NoData values. The pointer image is used to traverse the stream network and should only be created using the D8 algorithm. Background cells will be assigned the NoData value in the output image, unless the user specifies zero_background=True, in which case non-stream cells will be assigned zero values in the output.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, the user should specify esri_pntr=True.

Reference Shreve, R. L. (1966). Statistical law of stream numbers. The Journal

of Geology, 74(1), 17-37.

See Also

horton_stream_order, hack_stream_order, strahler_stream_order, topological_stream_order

Function Signature

def shreve_stream_magnitude(self, d8_pntr: Raster, streams_raster: Raster, esri_pntr: bool = False, zero_background: bool = False) -> Raster: ...

sigmoidal_contrast_stretch

This tool performs a sigmoidal stretch on a raster image. This is a transformation where the input image value for a grid cell (zin) is transformed to an output value zout such that:

zout = (1.0 / (1.0 + exp(gain(cutoff - z))) - a ) / b x num_tones

where,

z = (zin - MIN) / RANGE,

a = 1.0 / (1.0 + exp(gain x cutoff)),

b = 1.0 / (1.0 + exp(gain x (cutoff - 1.0))) - 1.0 / (1.0 + exp(gain x cutoff)),

MIN and RANGE are the minimum value and data range in the input image respectively and gain and cutoff are user specified parameters (gain, cutoff).

Like all of WhiteboxTools's contrast enhancement tools, this operation will work on either greyscale or RGB input images.

See Also

PiecewiseContrastStretch, gaussian_contrast_stretch, histogram_equalization, min_max_contrast_stretch, percentage_contrast_stretch, standard_deviation_contrast_stretch

Function Signature

def sigmoidal_contrast_stretch(self, raster: Raster, cutoff: float = 0.0, gain: float = 1.0, num_tones: int = 256) -> Raster: ...

singlepart_to_multipart

This tool can be used to convert a vector file containing single-part features into a vector containing multi-part features. The user has the option to either group features based on an ID Field (field flag), which is a categorical field within the vector's attribute table. The ID Field should either be of String (text) or Integer type. Fields containing decimal values are not good candidates for the ID Field. If no field flag is specified, all features will be grouped together into one large multi-part vector.

This tool works for vectors containing either point, line, or polygon features. Since vectors of a POINT VectorGeometryType cannot represent multi-part features, the VectorGeometryType of the output file will be modified to a MULTIPOINT VectorGeometryType if the input file is of a POINT VectorGeometryType. If the input vector is of a POLYGON VectorGeometryType, the user can optionally set the algorithm to search for polygons that should be represented as hole parts. In the case of grouping based on an ID Field, hole parts are polygon features contained within larger polygons of the same ID Field value. Please note that searching for polygon holes may significantly increase processing time for larger polygon coverages.

See Also

MultiPartToSinglePart

Function Signature

def singlepart_to_multipart(self, input: Vector, field_name: str) -> Vector: ...

sink

This tool measures the depth that each grid cell in an input (dem) raster digital elevation model (DEM) lies within a sink feature, i.e. a closed topographic depression. A sink, or depression, is a bowl-like landscape feature, which is characterized by interior drainage and groundwater recharge. The depth_in_sink tool operates by differencing a filled DEM, using the same depression filling method as fill_depressions, and the original surface model.

In addition to the names of the input DEM (dem) and the output raster (output), the user must specify whether the background value (i.e. the value assigned to grid cells that are not contained within sinks) should be set to 0.0 (zero_background) Without this optional parameter specified, the tool will use the NoData value as the background value.

Reference

Antonić, O., Hatic, D., & Pernar, R. (2001). DEM-based depth in sink as an environmental estimator. Ecological Modelling, 138(1-3), 247-254.

See Also

fill_depressions

Function Signature

def sink(self, dem: Raster, zero_background: bool = False) -> Raster: ...

slope

This tool calculates slope gradient (i.e. slope steepness in degrees, radians, or percent) for each grid cell in an input digital elevation model (DEM). The user must input a DEM (dem). The Z conversion factor is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z conversion factor.

The tool uses Horn's (1981) 3rd-order finite difference method to estimate slope. Given the following clock-type grid cell numbering scheme (Gallant and Wilson, 2000),

| 7 | 8 | 1 |
| 6 | 9 | 2 |
| 5 | 4 | 3 |

slope = arctan(fx2 + fy2)0.5

where,

fx = (z3 - z5 + 2(z2 - z6) + z1 - z7) / 8 * Δx

and,

fy = (z7 - z5 + 2(z8 - z4) + z1 - z3) / * Δy

Δx and Δy are the grid resolutions in the x and y direction respectively

Reference

Gallant, J. C., and J. P. Wilson, 2000, Primary topographic attributes, in Terrain Analysis: Principles and Applications, edited by J. P. Wilson and J. C. Gallant pp. 51-86, John Wiley, Hoboken, N.J.

See Also

aspect, plan_curvature, profile_curvature

Function Signature

def slope(self, dem: Raster, units: str = "degrees", z_factor: float = 1.0) -> Raster: ...

slope_vs_elev_plot

This tool can be used to create a slope versus average elevation plot for one or more digital elevation models (DEMs). Similar to a hypsometric analysis (hypsometric_analysis), the slope-elevation relation can reveal the basic topographic character of a site. The output of this analysis is an HTML document (output) that contains the slope-elevation chart. The tool can plot multiple slope-elevation analyses on the same chart by specifying multiple input DEM files (inputs). Each input DEM can have an optional watershed in which the slope-elevation analysis is confined by specifying the optional watershed flag. If multiple input DEMs are used, and a watershed is used to confine the analysis to a sub-area, there must be the same number of input raster watershed files as input DEM files. The order of the DEM and watershed files must the be same (i.e. the first DEM file must correspond to the first watershed file, the second DEM file to the second watershed file, etc.). Each watershed file may contain one or more watersheds, designated by unique identifiers.

See Also

hypsometric_analysis, SlopeVsAspectPlot

Function Signature

def slope_vs_elev_plot(self, dem_rasters: List[Raster], output_html_file: str, watershed_rasters: List[Raster]) -> None: ...

smooth_vectors

This tool smooths a vector coverage of either a POLYLINE or POLYGON base VectorGeometryType. The algorithm uses a simple moving average method for smoothing, where the size of the averaging window is specified by the user. The default filter size is 3 and can be any odd integer larger than or equal to 3. The larger the averaging window, the greater the degree of line smoothing.

Function Signature

def smooth_vectors(self, input: Vector, filter_size: int = 3) -> Vector: ...

snap_pour_points

This tool measures the depth that each grid cell in an input (dem) raster digital elevation model (DEM) lies within a sink feature, i.e. a closed topographic depression. A sink, or depression, is a bowl-like landscape feature, which is characterized by interior drainage and groundwater recharge. The depth_in_sink tool operates by differencing a filled DEM, using the same depression filling method as fill_depressions, and the original surface model.

In addition to the names of the input DEM (dem) and the output raster (output), the user must specify whether the background value (i.e. the value assigned to grid cells that are not contained within sinks) should be set to 0.0 (zero_background) Without this optional parameter specified, the tool will use the NoData value as the background value.

Reference

Antonić, O., Hatic, D., & Pernar, R. (2001). DEM-based depth in sink as an environmental estimator. Ecological Modelling, 138(1-3), 247-254.

See Also

fill_depressions

Function Signature

def snap_pour_points(self, pour_pts: Vector, flow_accum: Raster, snap_dist: float = 0.0) -> Vector: ...

sobel_filter

This tool performs a 3 × 3 or 5 × 5 Sobel edge-detection filter on a raster image. The Sobel filter is similar to the prewitt_filter, in that it identifies areas of high slope in the input image through the calculation of slopes in the x and y directions. The Sobel edge-detection filter, however, gives more weight to nearer cell values within the moving window, or kernel. For example, a 3 × 3 Sobel filter uses the following schemes to calculate x and y slopes:

X-direction slope

...
-101
-202
-101

Y-direction slope

...
121
000
-1-2-1

Each grid cell in the output image is assigned the square-root of the squared sum of the x and y slopes.

The user must specify the variant, including '3x3' and '5x5' variants. The user may also optionally clip the output image distribution tails by a specified amount (e.g. 1%).

See Also

prewitt_filter

Function Signature

def sobel_filter(self, raster: Raster, variant: str = "3x3", clip_tails: float = 0.0) -> Raster: ...

spherical_std_dev_of_normals

This tool can be used to calculate the spherical standard deviation of the distribution of surface normals for an input digital elevation model (DEM; dem). This is a measure of the angular dispersion of the surface normal vectors within a local neighbourhood of a specified size (filter). spherical_std_dev_of_normals is therefore a measure of surface shape complexity, texture, and roughness. The spherical standard deviation (s) is defined as:

s = √[-2ln(R / N)] × 180 / π

where R is the resultant vector length and N is the number of unit normal vectors within the local neighbourhood. s is measured in degrees and is zero for simple planes and increases infinitely with increasing surface complexity or roughness. Note that this formulation of the spherical standard deviation assumes an underlying wrapped normal distribution.

The local neighbourhood size (filter) must be any odd integer equal to or greater than three. Grohmann et al. (2010) found that vector dispersion, a related measure of angular dispersion, increases monotonically with scale. This is the result of the angular dispersion measure integrating (accumulating) all of the surface variance of smaller scales up to the test scale. A more interesting scale relation can therefore be estimated by isolating the amount of surface complexity associated with specific scale ranges. That is, at large spatial scales, s should reflect the texture of large-scale landforms rather than the accumulated complexity at all smaller scales, including microtopographic roughness. As such, this tool normalizes the surface complexity of scales that are smaller than the filter size by applying Gaussian blur (with a standard deviation of one-third the filter size) to the DEM prior to calculating R. In this way, the resulting distribution is able to isolate and highlight the surface shape complexity associated with landscape features of a similar scale to that of the filter size.

This tool makes extensive use of integral images (i.e. summed-area tables) and parallel processing to ensure computational efficiency. It may, however, require substantial memory resources when applied to larger DEMs.

References

Grohmann, C. H., Smith, M. J., & Riccomini, C. (2010). Multiscale analysis of topographic surface roughness in the Midland Valley, Scotland. IEEE Transactions on Geoscience and Remote Sensing, 49(4), 1200-1213.

Hodgson, M. E., and Gaile, G. L. (1999). A cartographic modeling approach for surface orientation-related applications. Photogrammetric Engineering and Remote Sensing, 65(1), 85-95.

Lindsay J. B., Newman* D. R., Francioni, A. 2019. Scale-optimized surface roughness for topographic analysis. Geosciences, 9(7) 322. DOI: 10.3390/geosciences9070322.

See Also

circular_variance_of_aspect, multiscale_roughness, edge_density, surface_area_ratio, ruggedness_index

Function Signature

def spherical_std_dev_of_normals(self, dem: Raster, filter_size: int = 11) -> Raster: ...

split_colour_composite

This tool can be used to split a red-green-blue (RGB) colour-composite image into three separate bands of multi-spectral imagery. The user must specify the input image (input) and output red, green, blue images.

See Also

create_colour_composite

Function Signature

def split_colour_composite(self, composite_image: Raster) -> Tuple[Raster, Raster, Raster]: ...

split_vector_lines

This tool can be used to divide longer vector lines (input) into segments of a maximum specified length (length).

See Also

AssessRoute

Function Signature

def split_vector_lines(self, input: Vector, segment_length: float) -> Vector: ...

split_with_lines

This tool splits the lines or polygons in one layer using the lines in another layer to define the breaking points. Intersection points between geometries in both layers are considered as split points. The input layer (input) can be of either POLYLINE or POLYGON VectorGeometryType and the output file will share this geometry type. The user must also specify an split layer (split), of POLYLINE VectorGeometryType, used to bisect the input geometries.

Each split geometry's attribute record will contain FID and PARENT_FID values and all of the attributes (excluding FID's) of the input layer.

See Also

'MergeLineSegments'

Function Signature

def split_with_lines(self, input: Vector, split_vector: Vector) -> Vector: ...

standard_deviation_contrast_stretch

This tool performs a standard deviation contrast stretch on a raster image. This operation maps each grid cell value in the input raster image (zin) onto a new scale that ranges from a lower-tail clip value (min_val) to the upper-tail clip value (max_val), with the user-specified number of tonal values (num_tones), such that:

zout = ((zin – min_val)/(max_val – min_val)) x num_tones

where zout is the output value. The values of min_val and max_val are determined based on the image mean and standard deviation. Specifically, the user must specify the number of standard deviations (clip or stdev) to be used in determining the min and max clip values. The tool will then calculate the input image mean and standard deviation and estimate the clip values from these statistics.

This is the same kind of stretch that is used to display raster type data on the fly in many GIS software packages.

See Also

PiecewiseContrastStretch, gaussian_contrast_stretch, histogram_equalization, min_max_contrast_stretch, percentage_contrast_stretch, sigmoidal_contrast_stretch

Function Signature

def standard_deviation_contrast_stretch(self, raster: Raster, clip: float = 2.0, num_tones: int = 256) -> Raster: ...

standard_deviation_filter

This tool performs a standard deviation filter on an input image (input). A standard deviation filter assigns to each cell in the output grid the standard deviation, a measure of dispersion, of the values contained within a moving window centred on each grid cell.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

See Also

range_filter, total_filter

Function Signature

def standard_deviation_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

standard_deviation_of_slope

Calculates the standard deviation of slope from an input DEM, a metric of roughness described by Grohmann et al., (2011).

Function Signature

def standard_deviation_of_slope(self, dem: Raster, filter_size: int = 11, z_factor: float = 1.0) -> Raster: ...

stochastic_depression_analysis

This tool performs a stochastic analysis of depressions within a DEM, calculating the probability of each cell belonging to a depression. This land-surface parameter (pdep) has been widely applied in wetland and bottom-land mapping applications.

This tool differs from the original Whitebox GAT tool in a few significant ways:

  1. The Whitebox GAT tool took an error histogram as an input. In practice people found it difficult to create this input. Usually they just generated a normal distribution in a spreadsheet using information about the DEM root-mean-square-error (RMSE). As such, this tool takes a RMSE input and generates the histogram internally. This is more convienent for most applications but loses the flexibility of specifying the error distribution more completely.

  2. The Whitebox GAT tool generated the error fields using the turning bands method. This tool generates a random Gaussian error field with no spatial autocorrelation and then applies local spatial averaging using a Gaussian filter (the size of which depends of the error autocorrelation length input) to increase the level of autocorrelation. We use the Fast Almost Gaussian Filter of Peter Kovesi (2010), which uses five repeat passes of a mean filter, based on an integral image. This filter method is highly efficient. This results in a significant performance increase compared with the original tool.

  3. Parts of the tool's workflow utilize parallel processing. However, the depression filling operation, which is the most time-consuming part of the workflow, is not parallelized.

In addition to the input DEM (dem) and output pdep file name (output), the user must specify the nature of the error model, including the root-mean-square error (rmse) and the error field correlation length (range, in map units). These parameters determine the statistical frequency distribution and spatial characteristics of the modeled error fields added to the DEM in each iteration of the simulation. The user must also specify the number of iterations (iterations). A larger number of iterations will produce a smoother pdep raster.

This tool creates several temporary rasters in memory and, as a result, is very memory hungry. This will necessarily limit the size of DEMs that can be processed on more memory-constrained systems. As a rough guide for usage, the computer system will need 6-10 times more memory than the file size of the DEM. If your computer possesses insufficient memory, you may consider splitting the input DEM apart into smaller tiles.

For a video demonstrating the application of the stochastic_depression_analysis tool, see this YouTube video.

Reference

Lindsay, J. B., & Creed, I. F. (2005). Sensitivity of digital landscapes to artifact depressions in remotely-sensed DEMs. Photogrammetric Engineering & Remote Sensing, 71(9), 1029-1036.

See Also

impoundment_size_index, fast_almost_gaussian_filter

Function Signature

def stochastic_depression_analysis(self, dem: Raster, rmse: float, range: float, iterations: int = 100) -> Raster: ...

strahler_order_basins

This tool will identify the catchment areas of each Horton-Strahler stream order link in a user-specified stream network (streams), i.e. the network's Strahler basins. The tool effectively performs a Horton-Strahler stream ordering operation (horton_stream_order) followed by by a watershed operation. The user must specify the name of a flow pointer (flow direction) raster (d8_pntr), a streams raster (streams), and the output raster (output). The flow pointer and streams rasters should be generated using the d8_pointer algorithm. This will require a depressionless DEM, processed using either the breach_depressions_least_cost or fill_depressions tool.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, the esri_pntr parameter must be specified.

NoData values in the input flow pointer raster are assigned NoData values in the output image.

See Also

horton_stream_order, watershed, d8_pointer, breach_depressions_least_cost, fill_depressions

Function Signature

def strahler_order_basins(self, d8_pointer: Raster, streams: Raster, esri_pntr: bool = False) -> Raster: ...

strahler_stream_order

This tool can be used to assign the Strahler stream order to each link in a stream network. Stream ordering is often used in hydro-geomorphic and ecological studies to quantify the relative size and importance of a stream segment to the overall river system. There are several competing stream ordering schemes. Based on to this common stream numbering system, headwater stream links are assigned an order of one. Stream order only increases downstream when two links of equal order join, otherwise the downstream link is assigned the larger of the two link orders.

Strahler order and Horton order are similar approaches to assigning stream network hierarchy. Horton stream order essentially starts with the Strahler order scheme, but subsequently replaces each of the assigned stream order value along the main trunk of the network with the order value of the outlet. The main channel is not treated differently compared with other tributaries in the Strahler ordering scheme.

The user must input a streams raster image (streams_raster) and D8 pointer (flow direction) image (d8_pntr). Stream cells are designated in the streams image as all positive, nonzero values. Thus all non-stream or background grid cells are commonly assigned either zeros or NoData values. The pointer image is used to traverse the stream network and should only be created using the D8 algorithm (d8_pointer). Background cells will be assigned the NoData value in the output image, unless the user specifies zero_background=True, in which case non-stream cells will be assigned zero values in the output.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, the user should specify esri_pntr=True.

Reference Strahler, A. N. (1957). Quantitative analysis of watershed

geomorphology. Eos, Transactions American Geophysical Union, 38(6), 913-920.

See Also

horton_stream_order, hack_stream_order, shreve_stream_magnitude, topological_stream_order

Function Signature

def strahler_stream_order(self, d8_pntr: Raster, streams_raster: Raster, esri_pntr: bool = False, zero_background: bool = False) -> Raster: ...

This tool identifies all interior and exterior links, and source, link, and sink nodes in an input stream network (streams_raster). The input streams raster is used to designate which grid cells contain a stream and the pointer image is used to traverse the stream network. Stream cells are designated in the streams image as all values greater than zero. Thus, all non-stream or background grid cells are commonly assigned either zeros or NoData values. Background cells will be assigned the NoData value in the output image, unless zero_background=True, in which case non-stream cells will be assigned zero values in the output.

Each feature is assigned the following identifier in the output image:

ValueStream Type
1Exterior Link
2Interior Link
3Source Node (head water)
4Link Node
5Sink Node

The user must input an input stream raster (streams_raster) and a pointer (flow direction) raster (d8_pntr). The flow pointer and streams rasters should be generated using the d8_pointer algorithm. This will require a depressionless DEM, processed using either the breach_depressions_least_cost or fill_depressions tools.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, set esri_pntr=True.

See Also

stream_link_identifier

Function Signature

def stream_link_class(self, d8_pntr: Raster, streams_raster: Raster, esri_pntr: bool = False, zero_background: bool = False) -> Raster: ...

This tool can be used to assign each link in a stream network a unique numeric identifier. This grid is used by a number of other stream network analysis tools.

The input streams raster (streams_raster) is used to designate which grid cells contain a stream and the pointer image is used to traverse the stream network. Stream cells are designated in the streams image as all values greater than zero. Thus, all non-stream or background grid cells are commonly assigned either zeros or NoData values. Background cells will be assigned the NoData value in the output image, unless the user specifies zero_background=True, in which case non-stream cells will be assigned zero values in the output.

The user must specify the name of a flow pointer (flow direction) raster (d8_pntr) and a streams raster (streams_raster). The flow pointer and streams rasters should be generated using the d8_pointer algorithm. This will require a depressionless DEM, processed using either the breach_depressions_least_cost or fill_depressions tool.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, set esri_pntr=True.

See Also

d8_pointer, tributary_identifier, breach_depressions_least_cost, fill_depressions

Function Signature

def stream_link_identifier(self, d8_pntr: Raster, streams_raster: Raster, esri_pntr: bool = False, zero_background: bool = False) -> Raster: ...

This tool can be used to measure the length of each link in a stream network. The user must input a stream link ID raster (streams_id_raster), created using the stream_link_identifier tool, and D8 pointer raster (d8_pointer). The flow pointer raster is used to traverse the stream network and should only be created using the d8_pointer algorithm. Stream cells are designated in the stream link ID raster as all non-zero, positive values. Background cells will be assigned the NoData value in the output image, unless zero_background=True, in which case non-stream cells will be assigned zero values in the output.

See Also

stream_link_identifier, d8_pointer, stream_link_slope

Function Signature

def stream_link_length(self, d8_pointer: Raster, streams_id_raster: Raster, esri_pointer: bool = False, zero_background: bool = False) -> Raster: ...

This tool can be used to measure the average slope gradient, in degrees, of each link in a raster stream network. To estimate the slope of individual grid cells in a raster stream network, use the stream_slope_continuous tool instead. The user must input a stream link identifier raster image (streams_id_raster), a D8 pointer image (d8_pointer), and a digital elevation model (dem). The pointer image is used to traverse the stream network and must only be created using the D8 algorithm (d8_pointer). Stream cells are designated in the streams image as all values greater than zero. Thus, all non-stream or background grid cells are commonly assigned either zeros or NoData values. Background cells will be assigned the NoData value in the output image, unless zero_background=True, in which case non-stream cells will be assigned zero values in the output.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, set esri_pointer=True.

See Also

stream_slope_continuous, d8_pointer

Function Signature

def stream_link_slope(self, d8_pointer: Raster, streams_id_raster: Raster, dem: Raster, esri_pointer: bool = False, zero_background: bool = False) -> Raster: ...

stream_slope_continuous

This tool can be used to measure the slope gradient, in degrees, each grid cell in a raster stream network. To estimate the average slope for each link in a stream network, use the stream_link_slope tool instead. The user must input a stream raster image (streams_raster), a D8 pointer image (d8_pointer), and a digital elevation model (dem). The pointer image is used to traverse the stream network and must only be created using the D8 algorithm (d8_pointer). Stream cells are designated in the streams image as all values greater than zero. Thus, all non-stream or background grid cells are commonly assigned either zeros or NoData values. Background cells will be assigned the NoData value in the output image, unless zero_background=True, in which case non-stream cells will be assigned zero values in the output.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, set esri_pointer=True.

See Also

stream_link_slope, d8_pointer

Function Signature

def stream_slope_continuous(self, d8_pointer: Raster, streams_raster: Raster, dem: Raster, esri_pointer: bool = False, zero_background: bool = False) -> Raster: ...

subbasins

This tool will identify the catchment areas to each link in a user-specified stream network, i.e. the network's sub-basins. subbasins effectively performs a stream link ID operation (stream_link_identifier) followed by a watershed operation. The user must specify the name of a flow pointer (flow direction) raster (d8_pntr), a streams raster (streams), and the output raster (output). The flow pointer and streams rasters should be generated using the d8_pointer algorithm. This will require a depressionless DEM, processed using either the breach_depressions_least_cost or fill_depressions tool.

hillslopes are conceptually similar to sub-basins, except that sub-basins do not distinguish between the right-bank and left-bank catchment areas of stream links. The Sub-basins tool simply assigns a unique identifier to each stream link in a stream network.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, the esri_pntr parameter must be specified.

NoData values in the input flow pointer raster are assigned NoData values in the output image.

See Also

stream_link_identifier, watershed, hillslopes, d8_pointer, breach_depressions_least_cost, fill_depressions

Function Signature

def subbasins(self, d8_pntr: Raster, streams: Raster, esri_pntr: bool = False) -> Raster: ...

sum_overlay

This tool calculates the sum for each grid cell from a group of raster images (inputs). NoData values in any of the input images will result in a NoData pixel in the output image (output).

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

weighted_sum, multiply_overlay

Function Signature

def sum_overlay(self, input_rasters: List[Raster]) -> Raster: ...

surface_area_ratio

This tool calculates the ratio between the surface area and planar area of grid cells within digital elevation models (DEMs). The tool uses the method of Jenness (2004) to estimate the surface area of a DEM grid cell based on the elevations contained within the 3 x 3 neighbourhood surrounding each cell. The surface area ratio has a lower bound of 1.0 for perfectly flat grid cells and is greater than 1.0 for other conditions. In particular, surface area ratio is a measure of neighbourhood surface shape complexity (texture) and elevation variability (local slope).

Reference

Jenness, J. S. (2004). Calculating landscape surface area from digital elevation models. Wildlife Society Bulletin, 32(3), 829-839.

See Also

ruggedness_index, multiscale_roughness, circular_variance_of_aspect, edge_density

Function Signature

def surface_area_ratio(self, dem: Raster) -> Raster: ...

symmetrical_difference

This tool will remove all the overlapping features, or parts of overlapping features, between input and overlay vector files, outputting only the features that occur in one of the two inputs but not both. The Symmetrical Difference is related to the Boolean exclusive-or (XOR) operation in set theory and is one of the common vector overlay operations in GIS. The user must specify the names of the input and overlay vector files as well as the output vector file name. The tool operates on vector points, lines, or polygon, but both the input and overlay files must contain the same VectorGeometryType.

The Symmetrical Difference can also be derived using a combination of other vector overlay operations, as either (A union B) difference (A intersect B), or (A difference B) union (B difference A).

The attributes of the two input vectors will be merged in the output attribute table. Fields that are duplicated between the inputs will share a single attribute in the output. Fields that only exist in one of the two inputs will be populated by null in the output table. Multipoint VectorGeometryTypes however will simply contain a single output feature identifier (FID) attribute. Also, note that depending on the VectorGeometryType (polylines and polygons), Measure and Z ShapeDimension data will not be transferred to the output geometries. If the input attribute table contains fields that measure the geometric properties of their associated features (e.g. length or area), these fields will not be updated to reflect changes in geometry shape and size resulting from the overlay operation.

See Also

intersect, difference, union, clip, erase

Function Signature

def symmetrical_difference(self, input: Vector, overlay: Vector, snap_tolerance: float = 2.220446049250313e-16) -> Vector: ...

tangential_curvature

This tool calculates the tangential curvature, which is the curvature of an inclined plan perpendicular to both the direction of flow and the surface (Gallant and Wilson, 2000). Curvature is a second derivative of the topographic surface defined by a digital elevation model (DEM). The user must input a DEM (dem). The output reports curvature in degrees multiplied by 100 for easier interpretation, as curvature values are often very small. The Z Conversion Factor (zfactor) is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z Conversion Factor. If the DEM is in the geographic coordinate system (latitude and longitude), with XY units measured in degrees, an appropriate Z Conversion Factor is calculated internally based on site latitude.

Reference

Gallant, J. C., and J. P. Wilson, 2000, Primary topographic attributes, in Terrain Analysis: Principles and Applications, edited by J. P. Wilson and J. C. Gallant pp. 51-86, John Wiley, Hoboken, N.J.

plan_curvature, profile_curvature, total_curvature, slope, aspect

Function Signature

def tangential_curvature(self, dem: Raster, log_transform: bool = False, z_factor: float = 1.0) -> Raster: ...

thicken_raster_line

This image processing tool can be used to thicken single-cell wide lines within a raster file along diagonal sections of the lines. Because of the limitation of the raster data format, single-cell wide raster lines can be traversed along diagonal sections without passing through a line grid cell. This causes problems for various raster analysis functions for which lines are intended to be barriers. This tool will thicken raster lines, such that it is impossible to cross a line without passing through a line grid cell. While this can also be achieved using a maximum filter, unlike the filter approach, this tool will result in the smallest possible thickening to achieve the desired result.

All non-zero, positive values are considered to be foreground pixels while all zero valued cells or NoData cells are considered background pixels.

Note: Unlike other filter-based operations in WhiteboxTools, this algorithm can't easily be parallelized because the output raster must be read and written to during the same loop.

See Also

line_thinning

Function Signature

def thicken_raster_line(self, raster: Raster) -> Raster: ...

time_in_daylight

This tool calculates the proportion of time a location is within daylight. That is, it calculates the proportion of time, during a user-defined time frame, that a grid cell in an input digital elevation model (dem) is outside of an area of shadow cast by a local object. The input DEM should truly be a digital surface model (DSM) that contains significant off-terrain objects. Such a model, for example, could be created using the first-return points of a LiDAR data set, or using the lidar_digital_surface_model tool.

The tool operates by calculating a solar almanac, which estimates the sun's position for the location, in latitude and longitude coordinate (lat, long), of the input DSM. The algorithm then calculates horizon angle (see horizon_angle) rasters from the DSM based on the user-specified azimuth fraction (az_fraction). For example, if an azimuth fraction of 15-degrees is specified, horizon angle rasters could be calculated for the solar azimuths 0, 15, 30, 45... In reality, horizon angle rasters are only calculated for azimuths for which the sun is above the horizon for some time during the tested time period. A horizon angle raster evaluates the vertical angle between each grid cell in a DSM and a distant obstacle (e.g. a mountain ridge, building, tree, etc.) that blocks the view along a specified direction. In calculating horizon angle, the user must specify the maximum search distance (max_dist) beyond which the query for higher, more distant objects will cease. This parameter strongly impacts the performance of the tool, with larger values resulting in significantly longer run-times. Users are advised to set the max_dist based on the maximum shadow length expected in an area. For example, in a relatively flat urban landscape, the tallest building will likely determine the longest shadow lengths. All grid cells for which the calculated solar positions throughout the time frame are higher than the cell's horizon angle are deemed to be illuminated during the time the sun is in the corresponding azimuth fraction.

By default, the tool calculates time-in-daylight for a time-frame spanning an entire year. That is, the solar almanac is calculated for each hour, at 10-second intervals, and for each day of the year. Users may alternatively restrict the time of year over which time-in-daylight is calculated by specifying a starting day (1-365; start_day) and ending day (1-365; end_day). Similarly, by specifying start time (start_time) and end time (end_time) parameters, the user is able to measure time-in-daylight for specific ranges of the day (e.g. for the morning or afternoon hours). These time parameters must be specified in 24-hour time (HH:MM:SS), e.g. 15:30:00. sunrise and sunset are also acceptable inputs for the start time and end time respectively. The timing of sunrise and sunset on each day in the tested time-frame will be determined using the solar almanac.

See Also

lidar_digital_surface_model, horizon_angle

Function Signature

def time_in_daylight(self, dem: Raster, az_fraction: float = 5.0, max_dist: float = float('inf'), latitude: float = 0.0, longitude: float = 0.0, utc_offset_str: str = "UTC+00:00", start_day: int = 1, end_day: int = 365, start_time: str = "sunrise", end_time: str = "sunset") -> Raster: ...

tin_interpolation

Creates a raster grid based on a triangular irregular network (TIN) fitted to vector points and linear interpolation within each triangular-shaped plane. The TIN creation algorithm is based on Delaunay triangulation.

The user must specify the attribute field containing point values (field). Alternatively, if the input Shapefile contains z-values, the interpolation may be based on these values (use_z). Either an output grid resolution (cell_size) must be specified or alternatively an existing base file (base) can be used to determine the output raster's (output) resolution and spatial extent. Natural neighbour interpolation generally produces a satisfactorily smooth surface within the region of data points but can produce spurious breaks in the surface outside of this region. Thus, it is recommended that the output surface be clipped to the convex hull of the input points (clip).

See Also

lidar_tin_gridding, construct_vector_tin, natural_neighbour_interpolation

Function Signature

def tin_interpolation(self, points: Vector, field_name: str = "FID", use_z: bool = False, cell_size: float = 0.0, base_raster: Raster = None, max_triangle_edge_length: float = float('inf')) -> Raster: ...

tophat_transform

This tool performs either a white or black top-hat transform on an input image. A top-hat transform is a common digital image processing operation used for various tasks, such as feature extraction, background equalization, and image enhancement. The size of the rectangular structuring element used in the filtering can be specified using the filterx and filtery flags.

There are two distinct types of top-hat transform including white and black top-hat transforms. The white top-hat transform is defined as the difference between the input image and its opening by some structuring element. An opening operation is the dilation (maximum filter) of an erosion (minimum filter) image. The black top-hat transform, by comparison, is defined as the difference between the closing and the input image. The user specifies which of the two flavours of top-hat transform the tool should perform by specifying either 'white' or 'black' with the variant flag.

See Also:

closing, opening, maximum_filter, minimum_filter

Function Signature

def tophat_transform(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11, variant: str = "white") -> Raster: ...

topological_stream_order

This tool can be used to assign the topological stream order to each link in a stream network. According to this stream numbering system, the link directly draining to the outlet is assigned an order of one. Each of the two tributaries draining to the order-one link are assigned an order of two, and so on until the most distant link from the catchment outlet has been assigned an order. The topological order can therefore be thought of as a measure of the topological distance of each link in the network to the catchment outlet and is likely to be related to travel time.

The user must input a streams raster image (streams_raster) and D8 pointer image (d8_pntr). Stream cells are designated in the streams image as all positive, nonzero values. Thus all non-stream or background grid cells are commonly assigned either zeros or NoData values. The pointer image is used to traverse the stream network and should only be created using the D8 algorithm. Background cells will be assigned the NoData value in the output image, unless the zero_background=True, in which case non-stream cells will be assigned zero values in the output.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, set esri_pntr=True.

See Also

hack_stream_order, horton_stream_order, strahler_stream_order, shreve_stream_magnitude

Function Signature

def topological_stream_order(self, d8_pntr: Raster, streams_raster: Raster, esri_pntr: bool = False, zero_background: bool = False) -> Raster: ...

total_curvature

This tool calculates the total curvature, which measures the curvature of the topographic surface rather than the curvature of a line across the surface in some direction (Gallant and Wilson, 2000). Total curvature can be positive or negative, with zero curvature indicating that the surface is either flat or the convexity in one direction is balanced by the concavity in another direction, as would occur at a saddle point. Curvature is a second derivative of the topographic surface defined by a digital elevation model (DEM). The user must input a DEM (dem).The output reports curvature in degrees multiplied by 100 for easier interpretation, as curvature values are often very small. The Z Conversion Factor (zfactor) is only important when the vertical and horizontal units are not the same in the DEM. When this is the case, the algorithm will multiply each elevation in the DEM by the Z Conversion Factor. If the DEM is in the geographic coordinate system (latitude and longitude), with XY units measured in degrees, an appropriate Z Conversion Factor is calculated internally based on site latitude.

Reference

Gallant, J. C., and J. P. Wilson, 2000, Primary topographic attributes, in Terrain Analysis: Principles and Applications, edited by J. P. Wilson and J. C. Gallant pp. 51-86, John Wiley, Hoboken, N.J.

plan_curvature, profile_curvature, tangential_curvature, slope, aspect

Function Signature

def total_curvature(self, dem: Raster, log_transform: bool = False, z_factor: float = 1.0) -> Raster: ...

total_filter

This tool performs a total filter on an input image. A total filter assigns to each cell in the output grid the total (sum) of all values in a moving window centred on each grid cell.

Neighbourhood size, or filter size, is specified in the x and y dimensions using the filterx and filtery flags. These dimensions should be odd, positive integer values (e.g. 3, 5, 7, 9, etc.).

See Also

range_filter

Function Signature

def total_filter(self, raster: Raster, filter_size_x: int = 11, filter_size_y: int = 11) -> Raster: ...

trace_downslope_flowpaths

This tool can be used to mark the flowpath initiated from user-specified locations downslope and terminating at either the grid's edge or a grid cell with undefined flow direction. The user must input the name of a D8 flow pointer grid (d8_pntr) and an input vector file indicating the location of one or more initiation points, i.e. 'seed points' (seed_pts). The seed point file must be a vector of the POINT VectorGeometryType. Note that the flow pointer should be generated from a DEM that has been processed to remove all topographic depression (see breach_depressions_least_cost and fill_depressions) and created using the D8 flow algorithm (d8_pointer).

See Also

d8_pointer, breach_depressions_least_cost, fill_depressions, downslope_flowpath_length, downslope_distance_to_stream

Function Signature

def trace_downslope_flowpaths(self, seed_points: Vector, d8_pointer: Raster, esri_pntr: bool = False, zero_background: bool = False) -> Raster: ...

travelling_salesman_problem

This tool finds approximate solutions to travelling salesman problems, the goal of which is to identify the shortest route connecting a set of locations. The tool uses an algorithm that applies a 2-opt heuristic and a 3-opt heuristic as a fall-back if the initial approach takes too long. The user must specify the names of the input points vector (input) and output lines vector file (output), as well as the duration, in seconds, over which the algorithm is allowed to search for improved solutions (duration). The tool works in parallel to find more optimal solutions.

Function Signature

def travelling_salesman_problem(self, input: Vector, duration: int = 60) -> Vector: ...

trend_surface

This tool can be used to interpolate a trend surface from a raster image. The technique uses a polynomial, least-squares regression analysis. The user must specify the name of the input raster file. In addition, the user must specify the polynomial order (1 to 10) for the analysis. A first-order polynomial is a planar surface with no curvature. As the polynomial order is increased, greater flexibility is allowed in the fitted surface. Although polynomial orders as high as 10 are accepted, numerical instability in the analysis often creates artifacts in trend surfaces of orders greater than 5. The operation will display a text report on completion, in addition to the output raster image. The report will list each of the coefficient values and the r-square value. Note that the entire raster image must be able to fit into computer memory, limiting the use of this tool to relatively small rasters. The Trend Surface (Vector Points) tool can be used instead if the input data is vector points contained in a shapefile.

Numerical stability is enhanced by transforming the x, y, z data by their minimum values before performing the regression analysis. These transform parameters are also reported in the output report.

Function Signature

def trend_surface(self, raster: Raster, output_html_file: str, polynomial_order: int = 1) -> Raster: ...

trend_surface_vector_points

This tool can be used to interpolate a trend surface from a vector points file. The technique uses a polynomial, least-squares regression analysis. The user must specify the name of the input shapefile, which must be of a 'Points' base VectorGeometryType and select the attribute in the shapefile's associated attribute table for which to base the trend surface analysis. The attribute must be numerical. In addition, the user must specify the polynomial order (1 to 10) for the analysis. A first-order polynomial is a planar surface with no curvature. As the polynomial order is increased, greater flexibility is allowed in the fitted surface. Although polynomial orders as high as 10 are accepted, numerical instability in the analysis often creates artifacts in trend surfaces of orders greater than 5. The operation will display a text report on completion, in addition to the output raster image. The report will list each of the coefficient values and the r-square value. The Trend Surface tool can be used instead if the input data is a raster image.

Numerical stability is enhanced by transforming the x, y, z data by their minimum values before performing the regression analysis. These transform parameters are also reported in the output report.

Function Signature

def trend_surface_vector_points(self, input: Vector, cell_size: float, output_html_file: str, field_name: str = "FID", polynomial_order: int = 1) -> Raster: ...

tributary_identifier

This tool can be used to assigns a unique identifier to each tributary in a stream network. A tributary is a section of a stream network extending from a channel head downstream to a confluence with a larger stream. Relative stream size is estimated using stream length as a surrogate. Tributaries therefore extend from channel heads downstream until a confluence is encountered in which the intersecting stream is longer, or an outlet cell is detected.

The input streams raster (streams_raster) is used to designate which grid cells contain a stream and the pointer image is used to traverse the stream network. Stream cells are designated in the streams image as all values greater than zero. Thus, all non-stream or background grid cells are commonly assigned either zeros or NoData values. Background cells will be assigned the NoData value in the output image, unless zero_background=True, in which case non-stream cells will be assigned zero values in the output.

The user must specify the name of a flow pointer (flow direction) raster (d8_pntr) and a streams raster (streams_raster). The flow pointer and streams rasters should be generated using the d8_pointer algorithm. This will require a depressionless DEM, processed using either the breach_depressions_least_cost or fill_depressions tool. flow direction) raster, and the output raster.

By default, the pointer raster is assumed to use the clockwise indexing method used by WhiteboxTools. If the pointer file contains ESRI flow direction values instead, set esri_pntr=True.

See Also

d8_pointer, stream_link_identifier, breach_depressions_least_cost, fill_depressions

Function Signature

def tributary_identifier(self, d8_pntr: Raster, streams_raster: Raster, esri_pntr: bool = False, zero_background: bool = False) -> Raster: ...

turning_bands_simulation

This tool can be used to create a random field using the turning bands algorithm. The user must specify the name of a base raster image (base) from which the output raster will derive its geographical information, dimensions (rows and columns), and other information. In addition, the range (range), in x-y units, must be specified. The range determines the correlation length of the resulting field. For a good description of how the algorithm works, see Carr (2002). The turning bands method creates a number of 1-D simulations (called bands) and fuses these together to create a 2-D error field. There is no natural stopping condition in this process, so the user must specify the number of bands to create (iterations). The default value of 1000 iterations is reasonable. The fewer iterations used, the more prevalent the 1-D simulations will be in the output error image, effectively creating artifacts. Run time increases with the number of iterations.

Turning bands simulation is a commonly applied technique in Monte Carlo style simulations of uncertainty. As such, it is frequently run many times during a simulation (often 1000s of times). When this is the case, algorithm performance and efficiency are key considerations. One alternative method to efficiently generate spatially autocorrelated random fields is to apply the fast_almost_gaussian_filter tool to the output of the random_field tool. This can be used to generate a random field with the desired spatial characteristics and frequency distribution. This is the alternative approach used by the stochastic_depression_analysis tool.

Reference

Carr, J. R. (2002). Data visualization in the geosciences. Upper Saddle River, NJ: Prentice Hall. pp. 267.

See Also

random_field, fast_almost_gaussian_filter, stochastic_depression_analysis

Function Signature

def turning_bands_simulation(self, base_raster: Raster = None, range: float = 1.0, iterations: int = 1000) -> Raster: ...

two_sample_ks_test

This tool will perform a two-sample Kolmogorov-Smirnov (K-S) test to evaluate whether a significant statistical difference exists between the frequency distributions of two rasters. The null hypothesis is that both samples come from a population with the same distribution. Note that this test evaluates the two input rasters for differences in their overall distribution shape, with no assumption of normality. If there is need to compare the per-pixel differences between two input rasters, a paired-samples test such as the paired_sample_t_test or the non-parametric wilcoxon_signed_rank_test should be used instead.

The user must specify the name of the two input raster images (input1 and input2) and the output report HTML file (output). The test can be performed optionally on the entire image or on a random sub-sample of pixel values of a user-specified size (num_samples). In evaluating the significance of the test, it is important to keep in mind that given a sufficiently large sample, extremely small and non-notable differences can be found to be statistically significant. Furthermore statistical significance says nothing about the practical significance of a difference.

See Also

KSTestForNormality, paired_sample_t_test, wilcoxon_signed_rank_test

Function Signature

def two_sample_ks_test(self, raster1: Raster, raster2: Raster, output_html_file: str, num_samples: int) -> None: ...

union

This tool splits vector layers at their overlaps, creating a layer containing all the portions from both input and overlay layers. The Union is related to the Boolean OR operation in set theory and is one of the common vector overlay operations in GIS. The user must specify the names of the input and overlay vector files as well as the output vector file name. The tool operates on vector points, lines, or polygon, but both the input and overlay files must contain the same VectorGeometryType.

The attributes of the two input vectors will be merged in the output attribute table. Fields that are duplicated between the inputs will share a single attribute in the output. Fields that only exist in one of the two inputs will be populated by null in the output table. Multipoint VectorGeometryTypes however will simply contain a single output feature identifier (FID) attribute. Also, note that depending on the VectorGeometryType (polylines and polygons), Measure and Z ShapeDimension data will not be transferred to the output geometries. If the input attribute table contains fields that measure the geometric properties of their associated features (e.g. length or area), these fields will not be updated to reflect changes in geometry shape and size resulting from the overlay operation.

See Also

intersect, difference, symmetrical_difference, clip, erase

Function Signature

def union(self, input: Vector, overlay: Vector, snap_tolerance: float = 2.220446049250313e-16) -> Vector: ...

unnest_basins

In some applications it is necessary to relate a measured variable for a group of hydrometric stations (e.g. characteristics of flow timing and duration or water chemistry) to some characteristics of each outlet's catchment (e.g. mean slope, area of wetlands, etc.). When the group of outlets are nested, i.e. some stations are located downstream of others, then performing a watershed operation will result in inappropriate watershed delineation. In particular, the delineated watersheds of each nested outlet will not include the catchment areas of upstream outlets. This creates a serious problem for this type of application.

The Unnest Basin tool can be used to perform a watershedding operation based on a group of specified pour points, i.e. outlets or target cells, such that each complete watershed is delineated. The user must specify the name of a flow pointer (flow direction) raster, a pour point raster, and the name of the output rasters. Multiple numbered outputs will be created, one for each nesting level. Pour point, or target, cells are denoted in the input pour-point image as any non-zero, non-NoData value. The flow pointer raster should be generated using the D8 algorithm.

Function Signature

def unnest_basins(self, d8_pointer: Raster, pour_points: Vector, esri_pntr: bool = False) -> List[Raster]: ...

unsharp_masking

Unsharp masking is an image edge-sharpening technique commonly applied in digital image processing. Admittedly, the name 'unsharp' seems somewhat counter-intuitive given the purpose of the filter, which is to enchance the definition of edge features within the input image (input). This name comes from the use of a blurred, or unsharpened, intermediate image (mask) in the process. The blurred image is combined with the positive (original) image, creating an image that exhibits enhanced feature definition. A caution is needed in that the output image, although clearer, may be a less accurate representation of the image's subject. The output may also contain more speckle than the input image.

In addition to the input (input) and output image files, the user must specify the values of three parameters: the standard deviation distance (sigma), which is a measure of the filter size in pixels, the amount (amount), a percentage value that controls the magnitude of each overshoot at edges, and lastly, the threshold (threshold), which controls the minimal brightness change that will be sharpened. Pixels with values differ after the calculation of the filter by less than the threshold are unmodified in the output image.

unsharp_masking works with both greyscale and red-green-blue (RGB) colour images. RGB images are decomposed into intensity-hue-saturation (IHS) and the filter is applied to the intensity channel. Importantly, the intensity values range from 0-1, which is important when setting the threshold value for colour images. NoData values in the input image are ignored during processing.

See Also

gaussian_filter, high_pass_filter

Function Signature

def unsharp_masking(self, raster: Raster, sigma: float = 0.75, amount: float = 100.0, threshold: float = 0.0) -> Raster: ...

update_nodata_cells

This tool will assign the NoData valued cells in an input raster (input1) the values contained in the corresponding grid cells in a second input raster (input2). This operation is sometimes necessary because most other overlay operations exclude areas of NoData values from the analysis. This tool can be used when there is need to update the values of a raster within these missing data areas.

See Also

IsNodata

Function Signature

def update_nodata_cells(self, input1: Raster, input2: Raster) -> Raster: ...

upslope_depression_storage

This tool estimates the average upslope depression storage depth using the FD8 flow algorithm. The input DEM (dem) need not be hydrologically corrected; the tool will internally map depression storage and resolve flowpaths using depression filling. This input elevation model should be of a fine resolution (< 2 m), and is ideally derived using LiDAR. The tool calculates the total upslope depth of depression storage, which is divided by the number of upslope cells in the final step of the process, yielding the average upslope depression depth. Roughened surfaces tend to have higher values compared with smoothed surfaces. Values, particularly on hillslopes, may be very small (< 0.01 m).

See Also

FD8FlowAccumulation, fill_depressions, depth_in_sink

Function Signature

def upslope_depression_storage(self, dem: Raster) -> Raster: ...

user_defined_weights_filter

NoData values in the input image are ignored during the convolution operation. This can lead to unexpected behavior at the edges of images (since the default behavior is to return NoData when addressing cells beyond the grid edge) and where the grid contains interior areas of NoData values. Normalization of kernel weights can be useful for handling the edge effects associated with interior areas of NoData values. When the normalization option is selected, the sum of the cell value-weight product is divided by the sum of the weights on a cell-by-cell basis. Therefore, if the kernel at a particular grid cell contains neighboring cells of NoData values, normalization effectively re-adjusts the weighting to account for the missing data values. Normalization also ensures that the output image will possess values within the range of the input image and allows the user to specify integer value weights in the kernel. However, note that this implies that the sum of weights should equal one. In some cases, alternative sums (e.g. zero) are more appropriate, and as such normalization should not be applied in these cases.

Function Signature

def user_defined_weights_filter(self, raster: Raster, weights: List[List[float]], kernel_center: str = "center", normalize_weights: bool = False) -> Raster: ...

vector_hex_binning

The practice of binning point data to form a type of 2D histogram, density plot, or what is sometimes called a heatmap, is quite useful as an alternative for the cartographic display of of very dense points sets. This is particularly the case when the points experience significant overlap at the displayed scale. The PointDensity tool can be used to perform binning based on a regular grid (raster output). This tool, by comparison, bases the binning on a hexagonal grid.

The tool is similar to the CreateHexagonalVectorGrid tool, however instead will create an output hexagonal grid in which each hexagonal cell possesses a COUNT attribute which specifies the number of points from an input points file (Shapefile vector) that are contained within the hexagonal cell.

In addition to the names of the input points file and the output Shapefile, the user must also specify the desired hexagon width (w), which is the distance between opposing sides of each hexagon. The size (s) each side of the hexagon can then be calculated as, s = w / [2 x cos(PI / 6)]. The area of each hexagon (A) is, A = 3s(w / 2). The user must also specify the orientation of the grid with options of horizontal (pointy side up) and vertical (flat side up).

See Also

LidarHexBinning, PointDensity, CreateHexagonalVectorGrid

Function Signature

def vector_hex_binning(self, vector_points: Vector, width: float, orientation: str = "h") -> Vector: ...

vector_lines_to_raster

This tool can be used to convert a vector lines or polygon file into a raster grid of lines. If a vector of one of the polygon VectorGeometryTypes is selected, the resulting raster will outline the polygons without filling these features. Use the VectorPolygonToRaster tool if you need to fill the polygon features.

The user must specify the name of the input vector (input) and the output raster file (output). The Field Name (field) is the field from the attributes table, from which the tool will retrieve the information to assign to grid cells in the output raster. Note that if this field contains numerical data with no decimals, the output raster data type will be INTEGER; if it contains decimals it will be of a FLOAT data type. The field must contain numerical data. If the user does not supply a Field Name parameter, each feature in the raster will be assigned the record number of the feature. The assignment operation determines how the situation of multiple points contained within the same grid cell is handled. The background value is the value that is assigned to grid cells in the output raster that do not correspond to the location of any points in the input vector. This value can be any numerical value (e.g. 0) or the string 'NoData', which is the default.

If the user optionally specifies the cell_size parameter then the coordinates will be determined by the input vector (i.e. the bounding box) and the specified Cell Size. This will also determine the number of rows and columns in the output raster. If the user instead specifies the optional base raster file parameter (base), the output raster's coordinates (i.e. north, south, east, west) and row and column count will be the same as the base file. If the user does not specify either of these two optional parameters, the tool will determine the cell size automatically as the maximum of the north-south extent (determined from the shapefile's bounding box) or the east-west extent divided by 500.

See Also

vector_points_to_raster, vector_polygons_to_raster

Function Signature

def vector_lines_to_raster(self, input: Vector, field_name: str = "FID", zero_background: bool = False, cell_size: float = 0.0, base_raster: Raster = None) -> Raster: ...

vector_points_to_raster

This tool can be used to convert a vector points file into a raster grid. The user must specify the name of the input vector and the output raster file. The field name (field) is the field from the attributes table from which the tool will retrieve the information to assign to grid cells in the output raster. The field must contain numerical data. If the user does not supply a field name parameter, each feature in the raster will be assigned the record number of the feature. The assignment operation determines how the situation of multiple points contained within the same grid cell is handled. The background value is zero by default but can be set to NoData optionally using the nodata value.

If the user optionally specifies the grid cell size parameter (cell_size) then the coordinates will be determined by the input vector (i.e. the bounding box) and the specified cell size. This will also determine the number of rows and columns in the output raster. If the user instead specifies the optional base raster file parameter (base), the output raster's coordinates (i.e. north, south, east, west) and row and column count will be the same as the base file.

In the case that multiple points are contained within a single grid cell, the output can be assigned (assign) the first, last (default), min, max, sum, or number of the contained points.

See Also

vector_polygons_to_raster, vector_lines_to_raster

Function Signature

def vector_points_to_raster(self, input: Vector, field_name: str = "FID", assign_op: str = "last", zero_background: bool = False, cell_size: float = 0.0, base_raster: Raster = None) -> Raster: ...

vector_polygons_to_raster

public constructor

Function Signature

def vector_polygons_to_raster(self, input: Vector, field_name: str = "FID", zero_background: bool = False, cell_size: float = 0.0, base_raster: Raster = None) -> Raster: ...

verbose

Determines whether tool functions output to stdout (wbe.verbose=True), or if output is suppressed (wbe.verbose=False).

version

Returns the Whitebox Workflows version information.

viewshed

This tool can be used to calculate the viewshed (i.e. the visible area) from a location (i.e. viewing station) or group of locations based on the topography defined by an input digital elevation model (DEM). The user must input a DEM (dem), a viewing station input vector file (stations) and the viewing height (height). Viewing station locations are specified as points within an input shapefile. The output image indicates the number of stations visible from each grid cell. The viewing height is in the same units as the elevations of the DEM and represent a height above the ground elevation from which the viewshed is calculated.

viewshed should be used when there are a relatively small number of target sites for which visibility needs to be assessed. If you need to assess general landscape visibility as a land-surface parameter, the visibility_index tool should be used instead.

Viewshed analysis is a very computationally intensive task. Depending on the size of the input DEM grid and the number of viewing stations, this operation may take considerable time to complete. Also, this implementation of the viewshed algorithm does not account for the curvature of the Earth. This should be accounted for if viewsheds are being calculated over very extensive areas.

See Also

visibility_index

Function Signature

def viewshed(self, dem: Raster, station_points: Vector, station_height: float = 2.0) -> Raster: ...

visibility_index

This tool can be used to calculate a measure of landscape visibility based on the topography of an input digital elevation model (DEM). The user must input DEM a (dem), the viewing height (height), and a resolution factor (res_factor). Viewsheds are calculated for a subset of grid cells in the DEM based on the resolution factor. The visibility index value (0.0-1.0) indicates the proportion of tested stations (determined by the resolution factor) that each cell is visible from. The viewing height is in the same units as the elevations of the DEM and represent a height above the ground elevation. Each tested grid cell's viewshed will be calculated in parallel. However, visibility index is one of the most computationally intensive geomorphometric indices to calculate. Depending on the size of the input DEM grid and the resolution factor, this operation may take considerable time to complete. If the task is too long-running, it is advisable to raise the resolution factor. A resolution factor of 2 will skip every second row and every second column (effectively evaluating the viewsheds of a quarter of the DEM's grid cells). Increasing this value decreases the number of calculated viewshed but will result in a lower accuracy estimate of overall visibility. In addition to the high computational costs of this index, the tool also requires substantial memory resources to operate. Each of these limitations should be considered before running this tool on a particular data set. This tool is best to apply on computer systems with high core-counts and plenty of memory.

See Also

viewshed

Function Signature

def visibility_index(self, dem: Raster, station_height: float = 2.0, resolution_factor: int = 8) -> Raster: ...

voronoi_diagram

This tool creates a vector Voronoi diagram for a set of vector points. The Voronoi diagram is the dual graph of the Delaunay triangulation. The tool operates by first constructing the Delaunay triangulation and then connecting the circumcenters of each triangle. Each Voronoi cell contains one point of the input vector points. All locations within the cell are nearer to the contained point than any other input point.

A dense frame of 'ghost' (hidden) points is inserted around the input point set to limit the spatial extent of the diagram. The frame is set back from the bounding box of the input points by 2 x the average point spacing. The polygons of these ghost points are not output, however, points that are situated along the edges of the data will have somewhat rounded (paraboloic) exterior boundaries as a result of this edge condition. If this property is unacceptable for application, clipping the Voronoi diagram to the convex hull may be a better alternative.

This tool works on vector input data only. If a Voronoi diagram is needed to tessellate regions associated with a set of raster points, use the euclidean_allocation tool instead. To use Voronoi diagrams for gridding data (i.e. raster interpolation), use the NearestNeighbourGridding tool.

See Also

construct_vector_tin, euclidean_allocation, NearestNeighbourGridding

Function Signature

def voronoi_diagram(self, input_points: Vector) -> Vector: ...

watershed

This tool will perform a watershedding operation based on a group of input vector pour points (pour_pts), i.e. outlets or points-of-interest. Watershedding is a procedure that identifies all of the cells upslope of a cell of interest (pour point) that are connected to the pour point by a flow-path. The user must input a D8-derived flow pointer (flow direction) raster (d8_pntr) and a vector pour point file (pour_pts). The pour points must be of a Point ShapeType (i.e. Point, PointZ, PointM, MultiPoint, MultiPointZ, MultiPointM). Watersheds will be assigned the input pour point FID value. The flow pointer raster must be generated using the D8 algorithm, d8_pointer.

Pour point vectors can be attained by on-screen digitizing to designate these points-of-interest locations. Because pour points are usually, although not always, situated on a stream network, it is recommended that you use Jenson's method (jenson_snap_pour_points) to snap pour points on the stream network. This will ensure that the digitized outlets are coincident with the digital stream contained within the DEM flowpaths. If this is not done prior to inputting a pour-point set to the watershed tool, anomalously small watersheds may be output, as pour points that fall off of the main flow path (even by one cell) in the D8 pointer will yield very different catchment areas.

If a raster pour point is specified instead of vector points, the watershed labels will derive their IDs from the grid cell values of all non-zero, non-NoData valued grid cells in the pour points file. Notice that this file can contain any integer data. For example, if a lakes raster, with each lake possessing a unique ID, is used as the pour points raster, the tool will map the watersheds draining to each of the input lake features. Similarly, a pour points raster may actually be a streams file, such as what is generated by the stream_link_identifier tool.

By default, the pointer raster is assumed to use the clockwise indexing method used by Whitebox Workflows. If the pointer file contains ESRI flow direction values instead, the esri_pntr must be True.

There are several tools that perform similar watershedding operations in Whitebox Workflows. watershed is appropriate to use when you have a set of specific locations for which you need to derive the watershed areas. Use the basins tool instead when you simply want to find the watersheds draining to each outlet situated along the edge of a DEM. The isobasins tool can be used to divide a landscape into roughly equally sized watersheds. The subbasins and strahler_order_basins are useful when you need to find the areas draining to each link within a stream network. Finally, hillslopes can be used to identify the areas draining the each of the left and right banks of a stream network.

Reference

Jenson, S. K. (1991), Applications of hydrological information automatically extracted from digital elevation models, Hydrological Processes, 5, 31–44, doi:10.1002/hyp.3360050104.

Lindsay JB, Rothwell JJ, and Davies H. 2008. Mapping outlet points used for watershed delineation onto DEM-derived stream networks, Water Resources Research, 44, W08442, doi:10.1029/2007WR006507.

See Also

d8_pointer, basins, subbasins, isobasins, strahler_order_basins, hillslopes, jenson_snap_pour_points, breach_depressions_least_cost, fill_depressions

Function Signature

def watershed(self, d8_pointer: Raster, pour_points: Vector, esri_pntr: bool = False) -> Raster: ...

watershed_from_raster_pour_points

This tool will perform a watershedding operation based on a group of input raster containing point points (pour_points). Watershedding is a procedure that identifies all of the cells upslope of a cell of interest (pour point) that are connected to the pour point by a flow-path. The user must input a D8-derived flow pointer (flow direction) raster (d8_pointer) and a pour points raster (pour_points). The flow pointer raster must be generated using the D8 algorithm, d8_pointer.

Watershed labels will derive their IDs from the grid cell values of all non-zero, non-NoData valued grid cells in the pour points file. Notice that this file can contain any integer data. For example, if a lakes raster, with each lake possessing a unique ID, is used as the pour points raster, the tool will map the watersheds draining to each of the input lake features. Similarly, a pour points raster may actually be a streams file, such as what is generated by the stream_link_identifier tool.

By default, the pointer raster is assumed to use the clockwise indexing method used by Whitebox Workflows. If the pointer file contains ESRI flow direction values instead, the esri_pntr parameter must be specified.

There are several tools that perform similar watershedding operations in Whitebox Workflows. watershed is appropriate to use when you have a set of specific locations for which you need to derive the watershed areas. Use the basins tool instead when you simply want to find the watersheds draining to each outlet situated along the edge of a DEM. The isobasins tool can be used to divide a landscape into roughly equally sized watersheds. The subbasins and strahler_order_basins are useful when you need to find the areas draining to each link within a stream network. Finally, hillslopes can be used to identify the areas draining the each of the left and right banks of a stream network.

Reference

Jenson, S. K. (1991), Applications of hydrological information automatically extracted from digital elevation models, Hydrological Processes, 5, 31–44, doi:10.1002/hyp.3360050104.

Lindsay JB, Rothwell JJ, and Davies H. 2008. Mapping outlet points used for watershed delineation onto DEM-derived stream networks, Water Resources Research, 44, W08442, doi:10.1029/2007WR006507.

See Also

d8_pointer, basins, subbasins, isobasins, strahler_order_basins, hillslopes, jenson_snap_pour_points, breach_depressions_least_cost, fill_depressions

Function Signature

def watershed_from_raster_pour_points(self, d8_pointer: Raster, pour_points: Raster, esri_pntr: bool = False) -> Raster: ...

weighted_overlay

This tool performs a weighted overlay on multiple input images. It can be used to combine multiple factors with varying levels of weight or relative importance. The WeightedOverlay tool is similar to the WeightedSum tool but is more powerful because it automatically converts the input factors to a common user-defined scale and allows the user to specify benefit factors and cost factors. A benefit factor is a factor for which higher values are more suitable. A cost factor is a factor for which higher values are less suitable. By default, WeightedOverlay assumes that input images are benefit factors, unless a cost value of 'true' is entered in the cost array. Constraints are absolute restriction with values of 0 (unsuitable) and 1 (suitable). This tool is particularly useful for performing multi-criteria evaluations (MCE).

Notice that the algorithm will convert the user-defined factor weights internally such that the sum of the weights is always equal to one. As such, the user can specify the relative weights as decimals, percentages, or relative weightings (e.g. slope is 2 times more important than elevation, in which case the weights may not sum to 1 or 100).

NoData valued grid cells in any of the input images will be assigned NoData values in the output image. The output raster is of the float data type and continuous data scale.

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

Function Signature

def weighted_overlay(self, factors: List[Raster], weights: List[float], cost: List[Raster] = None, constraints: List[Raster] = None, scale_max: float = 1.0) -> Raster: ...

weighted_sum

This tool performs a weighted-sum overlay on multiple input raster images. If you have a stack of rasters that you would like to sum, each with an equal weighting (1.0), then use the sum_overlay tool instead.

Warning

Each of the input rasters must have the same spatial extent and number of rows and columns.

See Also

sum_overlay

Function Signature

def weighted_sum(self, input_rasters: List[Raster], weights: List[float]) -> Raster: ...

wetness_index

This tool can be used to calculate the topographic wetness index, commonly used in the TOPMODEL rainfall-runoff framework. The index describes the propensity for a site to be saturated to the surface given its contributing area and local slope characteristics. It is calculated as:

WI = Ln(As / tan(Slope))

Where As is the specific catchment area (i.e. the upslope contributing area per unit contour length) estimated using one of the available flow accumulation algorithms in the Hydrological Analysis toolbox. Notice that As must not be log-transformed prior to being used; log-transformation of As is a common practice when visualizing the data. The slope image should be measured in degrees and can be created from the base digital elevation model (DEM) using the slope tool. Grid cells with a slope of zero will be assigned NoData in the output image to compensate for the fact that division by zero is infinity. These very flat sites likely coincide with the wettest parts of the landscape. The input images must have the same grid dimensions.

Grid cells possessing the NoData value in either of the input images are assigned NoData value in the output image. The output raster is of the float data type and continuous data scale.

See Also slope, D8FlowAccumulation, DInfFlowAccumulation, FD8FlowAccumulation, breach_depressions_least_cost

Function Signature

def wetness_index(self, specific_catchment_area: Raster, slope: Raster) -> Raster: ...

wilcoxon_signed_rank_test

This tool will perform a Wilcoxon signed-rank test to evaluate whether a significant statistical difference exists between the two rasters. The Wilcoxon signed-rank test is often used as a non-parametric equivalent to the paired-samples Student's t-test, and is used when the distribution of sample difference values between the paired inputs is non-Gaussian. The null hypothesis of this test is that difference between the sample pairs follow a symmetric distribution around zero. i.e. that the median difference between pairs of observations is zero.

The user must specify the name of the two input raster images (input1 and input2) and the output report HTML file (output). The test can be performed optionally on the entire image or on a random sub-sample of pixel values of a user-specified size (num_samples). In evaluating the significance of the test, it is important to keep in mind that given a sufficiently large sample, extremely small and non-notable differences can be found to be statistically significant. Furthermore statistical significance says nothing about the practical significance of a difference. Note that cells with a difference of zero are excluded from the ranking and tied difference values are assigned their average rank values.

See Also

paired_sample_test, two_sample_ks_test

Function Signature

def wilcoxon_signed_rank_test(self, raster1: Raster, raster2: Raster, output_html_file: str, num_samples: int) -> None: ...

working_directory

Returns the current working directory.

write_function_memory_insertion

Jensen (2015) describes write function memory (WFM) insertion as a simple yet effective method of visualizing land-cover change between two or three dates. WFM insertion may be used to qualitatively inspect change in any type of registered, multi-date imagery. The technique operates by creating a red-green-blue (RGB) colour composite image based on co-registered imagery from two or three dates. If two dates are input, the first date image will be put into the red channel, while the second date image will be put into both the green and blue channels. The result is an image where the areas of change are displayed as red (date 1 is brighter than date 2) and cyan (date 1 is darker than date 2), and areas of little change are represented in grey-tones. The larger the change in pixel brightness between dates, the more intense the resulting colour will be.

If images from three dates are input, the resulting composite can contain many distinct colours. Again, more intense the colours are indicative of areas of greater land-cover change among the dates, while areas of little change are represented in grey-tones. Interpreting the direction of change is more difficult when three dates are used. Note that for multi-spectral imagery, only one band from each date can be used for creating a WFM insertion image.

Reference

Jensen, J. R. (2015). Introductory Digital Image Processing: A Remote Sensing Perspective.

See Also

create_colour_composite, change_vector_analysis

Function Signature

def write_function_memory_insertion(self, image1: Raster, image2: Raster, image3: Raster) -> Raster: ...

write_lidar

Writes an in-memory Lidar object to disc.

Parameters

  • lidar: Lidar - An in-memory Lidar object
  • file_name: str - The name of the file on disc. If the file_name does not contain the full file path, the file will be written to the Whitebox working directory.

write_raster

Writes an in-memory Raster object to file.

Parameters

  • raster: Raster - The Raster object to write to disc.
  • file_name: str - The file name to write to. If the file name does not contain the full file path, the file will be written to the Whitebox working directory.
  • compress: bool - Boolean flag that determines whether the output file is compressed. Not all raster formats support compression. Default is False.

write_vector

Write an in-memory Vector object to disc.

Parameters

  • vector: Vector - The in-memory Vector object.
  • file_name: str - The file name to write to. If the file name does not contain the full file path, the file will be written to the Whitebox working directory.

z_scores

This tool will transform the values in an input raster image (input) into z-scores. Z-scores are also called standard scores, normal scores, or z-values. A z-score is a dimensionless quantity that is calculated by subtracting the mean from an individual raw value and then dividing the difference by the standard deviation. This conversion process is called standardizing or normalizing and the result is sometimes referred to as a standardized variable. The mean and standard deviation are estimated using all values in the input image except for NoData values. The input image should not have a Boolean or categorical data scale, i.e. it should be on a continuous scale.

See Also

cumulative_distribution

Function Signature

def z_scores(self, raster: Raster) -> Raster: ...

zonal_statistics

This tool can be used to extract common descriptive statistics associated with the distribution of some underlying data raster based on feature units defined by a feature definition raster. For example, this tool can be used to measure the maximum or average slope gradient (data image) for each of a group of watersheds (feature definitions). Although the data raster can contain any type of data, the feature definition raster must be categorical, i.e. it must define area entities using integer values.

The stat parameter can take the values, 'mean', 'median', 'minimum', 'maximum', 'range', 'standard deviation', or 'total'.

If an output image name is specified, the tool will assign the descriptive statistic value to each of the spatial entities defined in the feature definition raster. If text output is selected, an HTML table will be output, which can then be readily copied into a spreadsheet program for further analysis. This is a very powerful and useful tool for creating numerical summary data from spatial data which can then be interrogated using statistical analyses. At least one output type (image or text) must be specified for the tool to operate.

NoData values in either of the two input images are ignored during the calculation of the descriptive statistic.

See Also

raster_summary_stats

Function Signature

def zonal_statistics(self, data_raster: Raster, feature_definitions_raster: Raster, stat_type: str = "mean") -> Tuple[Raster, str]: ...