Raster Analysis

General raster analysis covers a broad set of cell-based operations: local algebra on single rasters, focal statistics across neighbourhoods, zonal summaries within polygon regions, reclassification, and suitability scoring. These tools form the computational backbone of many GIS modelling workflows.

This chapter demonstrates an environmental suitability analysis that combines multiple raster datasets through reclassification and weighted overlay.

Key Concepts

• Local operations: Applied to each cell independently — arithmetic, logic, trigonometry, conditional assignment. Result depends only on the cell's own value (and corresponding cells in other input rasters).
• Focal operations: Applied to each cell using values within a spatial neighbourhood (kernel). Examples: focal mean, focal max, focal standard deviation.
• Zonal statistics: Aggregate raster values by zone boundaries defined by a second raster or vector polygon layer.
• Reclassification: Map old cell values to new values via a lookup table or range intervals. Core step in suitability and habitat modelling.
• Weighted overlay: Combine multiple reclassified factor rasters using factor-specific weights. The weighted sum produces a composite suitability score.
• NoData / null handling: Cells with NoData propagate through most local operations. Ensure all input rasters share the same extent, resolution, and NoData mask before combining them.

End-to-End Workflow: Multi-Criteria Habitat Suitability

This workflow scores terrain for habitat suitability using slope, TWI, and distance from water as factors.

Inputs

All rasters must share the same projected CRS, extent, and cell size before being combined. Use Snap Raster Extents or QGIS Warp (Reproject) to align if needed.

Step 1 — Compute Distance from Water

Processing Toolbox → Whitebox Workflows → GIS Analysis → Euclidean Distance

Cells closer to water receive smaller values. This will be reclassified so that proximity = higher suitability.

Step 2 — Reclassify Slope Factor

Assign suitability scores 1–5 to slope ranges.

Processing Toolbox → Whitebox Workflows → Raster Analysis → Reclass

Use Reclass From File with a two-column table file (old value; new value) or set up intervals in the tool dialogue.

Step 3 — Reclassify TWI Factor

Processing Toolbox → Whitebox Workflows → Raster Analysis → Reclass
Input: twi.tif → Output: twi_reclass.tif

Step 4 — Reclassify Distance from Water

Closer = more suitable:

Processing Toolbox → Whitebox Workflows → Raster Analysis → Reclass
Input: dist_water.tif → Output: dist_water_reclass.tif

Step 5 — Weighted Overlay (Raster Calculator)

Combine the three factors using assigned weights (must sum to 1.0).

QGIS Raster Calculator:

("slope_reclass@1" * 0.4) + ("twi_reclass@1" * 0.35) + ("dist_water_reclass@1" * 0.25)

Output: suitability.tif (range 1–5, continuous).

Step 6 — Zonal Statistics (Optional)

Summarise suitability scores by catchment polygon.

Processing Toolbox → QGIS → Vector Analysis → Zonal Statistics (QGIS native)

Python Console Equivalent

import processing

# Step 1: distance from water
processing.run('whitebox_workflows:euclidean_distance', {
    'input': '/data/streams.tif',
    'output': '/data/dist_water.tif',
})

# Step 2–4: reclassify each factor
for src, dst in [
    ('slope', 'slope_reclass'),
    ('twi', 'twi_reclass'),
    ('dist_water', 'dist_water_reclass'),
]:
    processing.run('whitebox_workflows:reclass_from_file', {
        'input': f'/data/{src}.tif',
        'reclass_vals': f'/data/{src}_reclass.txt',
        'output': f'/data/{dst}.tif',
    })

# Step 5: weighted overlay via Raster Calculator
processing.run('qgis:rastercalculator', {
    'EXPRESSION': '("slope_reclass@1" * 0.4) + ("twi_reclass@1" * 0.35) + ("dist_water_reclass@1" * 0.25)',
    'LAYERS': [
        '/data/slope_reclass.tif',
        '/data/twi_reclass.tif',
        '/data/dist_water_reclass.tif',
    ],
    'OUTPUT': '/data/suitability.tif',
})

print("Suitability analysis complete.")

Advanced: Focal Statistics

Focal statistics smooth or enhance spatial patterns at a neighbourhood scale.

Processing Toolbox → Whitebox Workflows → Raster Analysis → Mean Filter

Use Standard Deviation Filter to highlight areas of high local variability, or Percentile Filter for rank-based neighbourhood smoothing.

Common Pitfalls

Validation Checklist

• All input rasters aligned (same CRS, extent, cell size, NoData value).
• Reclass table covers the full observed value range with no gaps.
• Weighted overlay weights sum to 1.0.
• Output suitability range matches expected 1–5 interval.
• Zonal statistics polygon CRS matches raster CRS.
• Focal filter kernel size is appropriate for target feature scale.