Learning Objectives

• List the data analysis operations available in GIS
• Explain how features are measured in GIS
• Describe methods of querying spatial and attribute data
• Provide examples of network analysis
• Outline methods of surface analysis
• Describe the differences between analysis carried out on raster and vector data
• Predict some of the problems that might be faced when conducting data analysis with GIS

Example questions for GIS analysis

• Location
  • Where is ...?
• Patterns
  • Land use, traffic flows ...
• Trends
  • Changes over time, population growth ...
• Conditions
  • Find a ... which fulfils certain conditions ...
• Implications
  • If we ... what's the impact ...?

Spatial Query(PostGIS examples)

select a.saps_label from localauthorities a, localauthorities b 
    where st_touches(a.the_geom, b.the_geom) 
    and b.saps_label like '%Kerry%';
 
select astext(the_geom), classdesc from archeology 
    where classdesc like '%Castle%';

Map Overlay

• The ability to integrate data from two sources
• Visual comparison
  • Two or more layers overlaid – no new data produced
  • Merge data from two or more layers – new data produced
• Vector and raster overlay
  • Vector
    • Complex & time-consuming
    • Relies on geometry & topology
  • Raster
    • Quick and efficient

• Point-in-polygon
• Line-in-polygon

• Polygon-on-polygon

• Intersections of lines and polygons need to be calculated using geometry

• Overlay of two or more layers results in more complex output layer

Raster Overlay

• Point = cell
• Line = string of cells

• Polygon = group of cells

• Map algebra
  • Input data layers can be added, subtracted, multiplied and divided to produce output data
  • Mathematical operations performed on individual cells from two or more input layers to produce an output value

Issues affecting vector and raster overlay

• Resolution
• Scale
  • Take care when disaggregating data
  • Take care not to mix nominal, ordinal, interval, ratio, cyclic data inappropriately
• Modifiable areal unit problem (MAUP)
  • Boundaries arbitrarily defined – may not reflect breaks in data
  • Internal homogeneity may be assumed but not exist
• Ecological fallacy
  • Applying characteristics of aggregated data to individuals

• Selecting threshold criteria
  • Define criteria for computations - “garbage in, garbage out”
• Visual complexity
  • Output may be difficult to interpret # Spatial interpolation
• Estimate the values of properties at unsampled sites within an area covered by existing observations
• Fill in the gaps
• Observations can be
  • Stratified
  • Clustered
  • Random
• Output subject to degree of uncertainty
• Common application: generate height contours

Surface analysis

• Draping
• Slope / aspect calculation
• Visibility analysis
• Analytical hillshade anallysis
• Curvature calculations

Slope and Aspect

• Calculated from a grid of elevations (a digital elevation model)
• Slope and aspect are calculated at each point in the grid, by comparing the point’s elevation to that of its neighbours usually its eight neighbours but the exact method varies in a scientific study, it is important to know exactly what method is used when calculating slope, and exactly how slope is defined

Slope Definitions

• Slope defined as an angle
  • … or rise over horizontal run
  • … or rise over actual run
• various methods
  • important to know how your favourite GIS calculates slope

Slope and Aspect

• measured from an elevation or bathymetry raster
• compare elevations of points in a 3x3 neighbourhood
• slope and aspect at one point estimated from elevations of it and surrounding 8 points
• number points row by row, from top left from 1 to 9

Typical Slope Calculation

• tan (slope) = sqrt (b2 + c2)
• b = (z3 + 2z6 + z9 - z1 - 2z4 - z7) / 8D

• c = (z1 + 2z2 + z3 - z7 - 2z8 - z9) / 8D

• b denotes slope in the x direction
• c denotes slope in the y direction

• D is the spacing of points (30 m)

• find the slope that fits best to the 9 elevations
• minimizes the total of squared differences between point elevation and the fitted slope

• weighting four closer neighbours higher

Aspect

• tan (aspect) = b/c
• Angle between vertical and direction of steepest slope
• Measured clockwise
• add 180 to aspect if c is positive, 360 to aspect if c is negative and b is positive