Lecture 9 - Raster Analysis Flashcards
Raster Analysis Processes (10)
- Feature to raster conversion
- Reclassification
- Raster to feature conversion
- Interpolate to raster
- Create contours
- Neighbourhood statistics
- Zonal statistics
- 3D analyst profile
- Zonal histogram
- Cut and fill
Feature to raster conversion
Conversion from feature
- ASCII to raster (numeric file)
- DEM to raster (different standards from countries)
- point to raster
- polygon to raster
- polyline to raster
Feature to raster (points)
Raster cell will have the value of the type of point in that cell. If there are multiple types of points it will by default use the first point it finds. You can also choose different parameters, like most frequent or select a priority. You can also do standard deviation or other. calculations
Feature to raster (lines)
Raster cell value can be based on the line with:
- the longest length through the cell (max length)
- the longest combined length
- priority ranking
Issue: pixelization because of resolution of cells
Reclassification
Reclassifies or changes cell values/cell group values to alternative values using a variety of methods
- specified intervals (group values into 10 intervals)
- by area (group into 10 groups containing same number of cells)
Use: replace cell values to study changes over time, simplify /group info
Example: suitability for new school from 205
Raster to feature conversion
converts raster values to vector feature
-polygon/point/line
Quantised output
-based on cell size
Issue: pixelization because of resolution of cells, for points, it will be in the middle of the pixel
Interpolation
Converts point positions and values into a continuous raster surface
- interpolates between points
- interpolates values, not density (different to KDE)
- a variety of different methods which give different results (Kriging, Spline, Inverse Distance Weighting - IDW)
Inverse Distance Weighting
Estimates cell values by averaging values of sample data points in neighbourhood of each processing cell. The closer a point to centre of estimated cell, the more influence (weight), it has in averaging process.
Kriging
Advanced geostatistical procedure generating estimated surface from scattered set of points with z-values. A thorough investigation of spatial behaviour of phenomena. Assumes distance/direction between points reflects spatial autocorrelation.
Spline
Uses mathematical function that minimises overall surface curvature, resulting in smooth raster surface passing exactly through input points
Create contours
Takes a raster surface and creates contour lines (lines of equal value)
- used for elevation
- also used for concentration
- can be set to any interval
- metres/degrees/etc
Neighbourhood statistics
Series of tools to analyse cells surrounding features of interest
- block statistics (i.e. to decrease raster resolution): non-overlapping neighbours
- focal statistics (overlapping neighbours)
- point statistics
- line statistics
Zonal statistics
Creates a raster surface of values from one raster/statistics within discrete zones from another raster
- can be used with input vector zones too
- zones = areas of interest
Zonal histogram
Creates a histogram of raster values within discrete zones
- like zonal statistics
- can use raster or feature ‘zones’ as input
- recommend use with raster
- gives frequency of each thing within particular zones so we can compare areas or change over time
3d analyst tool
Creates profiles (cross-sections) of elevation, can also be used on any type of raster to plot changes in value over distance like concentration or climate
