1 Levels of Measurement
- Nominal: Categories with no order (land use: forest, water, urban).
- Ordinal: Ranked categories (road class: highway > secondary > track).
- Interval: Equal intervals, no true zero (temperature in °C).
- Ratio: Equal intervals + true zero (elevation, population, area).
2 Vector vs Raster Data Models
| Aspect | Vector | Raster |
|---|---|---|
| Structure | Points, lines, polygons | Grid of cells (pixels) |
| Storage | Compact for sparse data | Large (every cell stored) |
| Topology | Supported | Not inherent |
| Best for | Discrete features (roads, parcels) | Continuous phenomena (elevation, temperature) |
| Overlay analysis | Complex (intersection) | Simple (cell-by-cell) |
| Accuracy | High (coordinate-based) | Depends on cell size |
3 Raster Compression Techniques
- Run-Length Encoding (RLE): Store (value, count) pairs. Efficient for simple rasters with long runs of same value.
- Quadtree: Recursively divide raster into quadrants until each is homogeneous. Good for mixed-complexity data.
- Block Coding: Group identical adjacent cells into rectangular blocks.
- Chain Coding: Encode boundaries as direction codes (Freeman chain code).
4 Vector Labelling and Feature Types
- Point: Single coordinate pair. Examples: well, GPS station, school.
- Line (arc): Ordered set of coordinates. Examples: road, river, contour.
- Polygon: Closed area. Examples: district, lake, land parcel.
- Labels link features to attribute tables via unique IDs.
5 Topology
Topological data model: Node → Arc → Polygon. Left/right polygon tables. Enables adjacency/containment queries without coordinate computation.
6 TIN (Triangulated Irregular Network)
- Vector-based surface model using irregular triangles (Delaunay triangulation).
- Dense triangles in complex terrain, sparse in flat areas.
- Efficient for surface representation; used for slope, aspect, volume calculations.
7 BSQ, BIL, BIP (Raster Storage Formats)
- BSQ (Band Sequential): All pixels of band 1, then band 2, etc. Best for single-band access.
- BIL (Band Interleaved by Line): Line 1 of all bands, then line 2, etc. Best for spatial access.
- BIP (Band Interleaved by Pixel): All bands of pixel 1, then pixel 2, etc. Best for spectral access.
Use comparison table with 6+ aspects. Draw simple diagrams of each. Mention advantages/disadvantages.
RLE: show a grid and its encoded form. Quadtree: show recursive subdivision. Block coding and chain coding briefly.
Nominal, ordinal, interval, ratio. Give one GIS example per level. Explain which statistical operations apply to each.
Connectivity, area definition, contiguity. Node-arc-polygon model. Enables spatial queries: adjacency, containment, network routing.
Delaunay triangulation from sample points. Variable resolution. Used for surface analysis: slope, aspect, volume, visibility.
Draw memory layout diagrams for each. BSQ = band access. BIL = spatial access. BIP = spectral access.