Ctrl+K

Introduction of Spatial Analysis

7. Introduction of Spatial Analysis#

7.1. What’s spatial analysis#

  1. Transformation, manipulation and application of analytical methods to spatial (geographic) data (Goodchild)

  2. It’s about geospatial kdd (Knowledge discovery from data)

    data→information→knowledge→wisdom

7.1.1. What’s “geospatial”#

  • Location + value (attribute)

    • When location changes, the information content of data changes

  • non-spatial: location does not matter → spatial invariance

7.1.2. Components (workflow) of spatial data analysis#

  1. Mapping and geo-visualization

    • Showing interesting patterns

  2. Exploratory spatial data analysis

    • Discovering interesting patterns

  3. Spatial modeling

    • Explaining interesting patterns

    • Optimization, Simulation, Prediction

7.2. What’s Spatial Questions#

  1. Where do things happen

    • Pattern, clusters, hot spots, disparities

  2. Why do they happen

    • Location decisions

    • Spatial regression, correlation analysis

  3. How does where things happen affect other things

    • (context, environment)

    • Spatial autocorrelation

  4. How does context affect what happens

    • Interaction

  5. Where should things be located

    • optimization

7.3. Spatial data types#

Remember match between the scale of process you study and the scale of measurement

7.3.1. Point#

  1. Characteristics

    • Location of events

  2. Research question

    • Randomly in space or clustered?

    • Point pattern analysis

      1. KDE(Heat Map)

7.3.2. Surface#

  1. Characteristics

    • Continuous spatial field

      • Air quality, noise, price

    • Research question

      1. Given discrete measures, what’s an air quality surface for a region

      2. Spatial interpolation

        1. Kriging

7.3.3. Discrete spatial data - lattice data#

  1. Characteristics

    • Area units

      • Census tracts, counties, countries

    • Research question

      • Where are hot spots of income inequality in the city

        1. Cluster detection

        2. Location similarity + Attribute similarity

7.3.4. Networks#

7.4. Modifiable Areal Unit Problem (MAUP)#

  1. What’s the proper spatial scale of analysis?

  2. Spatial heterogeneity - different processes at different locations/scales

  3. Both size and spatial arrangement will affect the result

    image.png

7.5. Change of Support Problem#

  1. Variables measured at different spatial scales

    • Nested, hierarchical structures (county / states)

    • Non-nested, overlapping (school district / census tract)

  2. Solution

    • Aggregate up to a common scale

    • Interpolate/impute - Bayesian approach

Contents