1 Spatial Interpolation
Estimating unknown values at unsampled locations from known sample points.
- IDW (Inverse Distance Weighted): Closer points have more influence. Weight = 1/dp (p usually 2). Deterministic, simple, fast; no error estimate.
- Kriging: Geostatistical method using semivariograms; considers spatial autocorrelation; provides error map; more accurate but slower.
- Splines: Fits smooth mathematical surface through points; can overshoot; exact interpolator.
2 Raster Analysis Functions
- Map Algebra: Cell-by-cell arithmetic on rasters.
- Slope: Rate of elevation change; output in degrees or percent.
- Aspect: Direction of steepest descent (0–360°).
- Hillshade: Simulated illumination for 3D visualisation.
- Viewshed: Binary raster of visible areas from an observer point.
- Euclidean Distance: Distance from each cell to nearest source.
- Reclassify: Reassign cell values to new categories.
- Zonal Statistics: Statistics for zones defined by another raster.
3 Local, Focal, and Zonal Operations
- Local: Output depends only on same-position cells. Example: NDVI.
- Focal/Neighbourhood: Output depends on surrounding cells (3×3 window). Example: slope, aspect.
- Zonal: Output computed for groups of cells (zones). Example: mean elevation per land-use class.
4 Weighted Overlay (Suitability Analysis)
- Reclassify all input rasters to a common suitability scale (e.g., 1–10).
- Assign weights to each factor (must sum to 100%).
- Score = Σ(Reclassified_value × Weight).
- Apply threshold to produce final suitability map.
5 NDVI
Range: −1 to +1. <0 = water, 0–0.1 = bare soil, 0.2–0.4 = sparse, 0.5–0.8 = dense vegetation.
For 10+ mark questions, explain at least 5 functions with one example each. Include sample input/output grids where possible.
Define each. IDW: deterministic, simple. Kriging: geostatistical, error estimate. Splines: smooth, overshoots. Compare in a table.
Values 160, 170, 180, 90 at distances 1000, 400, 600, 120 m. W = 1/d², Z = Σ(Z×W)/ΣW. Result approx 99.7.
Reclassify slope, soil, land use to 1–10. Apply weights (Slope 50%, Soil 30%, LU 20%). Compute cell-by-cell. Threshold ≥ 6 = suitable.
NDVI uses only same-position pixels in NIR and Red (local). Slope/aspect use a 3×3 window of surrounding elevation values (focal).
Slope from DEM → reclassify. Euclidean distance from rivers → reclassify. Reclassify land cover and soil. Map Algebra: multiply all binary layers.