10: Analysing Raster Data

Question 1

SPATIAL ANALYSIS PURPOSE

Answer 1

turns raw spatial data into useful geographic information

reveals spatial patterns, trends, anomalies, etc which might otherwise be missed

provides a systematic check on human intuitions, use scientific evidence for decision making (evidence based policy)

examples: where to put a new road, how to site something, etc

Question 2

BASIC ANALYSES METHODS OF RASTER GEOMETRY

Answer 2

Elementary operations
Macro-operations
Filtering methods

Question 3

EXPLAIN HOW RASTER OVERLAY WORKS

Answer 3

overlay of two polygon rasters with the same footprint

results in new polygon raster layer with all intersections of input raster layers

each polygon now contains both attributes from the two input layers

Question 4

SITE SUITABILITY EXAMPLE

Answer 4

work out where land is suitable for growing crops and need to identify areas for new farms

what kind of information is needed:

• soil productivity*
• climate
• land use/cover (vacant land)*
• accessibility to markets/demand (population)*
• accessibility to roads*
• water supply
• topography
• crop suitability
• protected areas/conservation*

what do we do with this data:

• soil productivity: reclassify in binary ‘suitable soils’
• land cover: reclassify in a binary way in ‘suitable areas’
• population: clip to identify areas with high population
• roads: buffer the major roads, perhaps 2km
• protected areas: reclassify in a binary way in ‘not protected areas’

combine with weighted overlay:
50% soil classification
20% distance to population centres
20% land cover/land use
10% distance to major roads
0% conservation

Question 5

Elementary Operations (3)

Answer 5

Radiometric transformation

• apply transfer function to grey values of all cells in a raster
• used to enhance visual discriminability
• linear function y=2x (duplicate grey values to spread spectrum / enhance contrast)

Thesholding

• set all grey values below a set threshold to 0; above the threshold to a constant value
• generates a binary raster image

Slicing

• suppress all values below or above a certain interval
• used to colour all zones between a certain range

Multi-criteria map query

Reclassify

Translation
-shift pattern for a given distance (depending on a metric)

Arithmetic & logical combinations

• starting with a binary map
• logical TRUE, FALSE (1,0) corresponding to 2 possible grey values per cell

Question 6

Elementary Operations

Answer 6

Radiometric transformation
Thesholding
Slicing
Multi-criteria map query
Reclassify
Translation
Arithmetic & logical combinations

Question 7

Logical combinations

Answer 7

AND
OR
NOR (exclusive OR)

Question 8

Computing buffers

Answer 8

create a new object consisting of areas within a user-defined distance of an existing object

can be done with raster or vector data structure

Question 9

Arithmetic & logical combinations

Answer 9

starting with a binary map

logical TRUE, FALSE (1,0) corresponding to 2 possible grey values per cell

Question 10

Translation

Answer 10

shift pattern for a given distance (depending on a metric)

Question 11

Slicing

Answer 11

• suppress all values below or above a certain interval

- used to colour all zones between a certain range

Question 12

Thresholding

Answer 12

• set all grey values below a set threshold to 0; above the threshold to a constant value
• generates a binary raster image

Question 13

Radiometric Transformation

Answer 13

• apply transfer function to grey values of all cells in a raster
• used to enhance visual discriminability
• linear function y=2x (duplicate grey values to spread spectrum / enhance contrast)

Question 14

Site suitability example methodology

Answer 14

1. collected data layers - raster if available
2. merged soil classification data tiles (by county) into a single file
3. converted vector format files to raster
4. resampled all 30m cell size raster files to 90m
5. use spatial analysis to create a kernel density map from census block population data
6. use spatial analysis to create a distance from major roads map
7. reclassed all input layers into integer values on a scale of 1-9
8. performed a weighted overlay raster analysis