| Geospatial Analysis - a comprehensive guide. 3rd edition © 2006-2009 de Smith, Goodchild, Longley |
| Contents, figures and tables |
|
|
Copyright and licensing information
1 Introduction and terminology
1.1.2 Spatial analysis, GIS and software tools
1.1.3 Intended audience and scope
1.2 Software tools and Companion Materials
1.2.1 GIS and related software tools
1.2.1.1 Sample software products
1.3 Terminology and Abbreviations
1.4 Common Measures and Notation
1.4.2 Statistical measures and related formulas
1.4.2.1 Counts and specific values
1.4.2.2 Measures of centrality
1.4.2.4 Measures of distribution shape
1.4.2.5 Measures of complexity and dimensionality
1.4.2.7 Data transforms and back transforms
2 Conceptual Frameworks for Spatial Analysis
2.1 The geospatial perspective
2.2.5 Multiple properties of places
2.2.9 Detail, resolution, and scale
2.3.2 Distance, direction and spatial weights matrices
2.3.3 Multidimensional scaling
2.3.10 Smoothing and sharpening
2.3.11 First- and second-order processes
2.5 Spatial Data Infrastructure
3.1 Spatial analysis as a process
3.3 Spatial analysis and the PPDAC model
3.3.1 Problem: Framing the question
3.3.2 Plan: Formulating the approach
3.3.4 Analysis: Analytical methods and tools
3.3.5 Conclusions: Delivering the results
3.4 Geospatial analysis and model building
3.5 The changing context of GIScience
4 Building Blocks of Spatial Analysis
4.1 Spatial Data Models and Methods
4.2 Geometric and Related Operations
4.2.1 Length and area for vector data
4.2.2 Length and area for raster datasets
4.2.3.2 Terrestrial (unprojected) surface area
4.2.4 Line Smoothing and point-weeding
4.2.5.1 Polygon centroids and centers
4.2.6 Point (object) in polygon (PIP)
4.2.9 Overlay and combination operations
4.2.11 Districting and re-districting
4.2.12 Classification and clustering
4.2.12.1 Univariate classification schemes
4.2.12.2 Multivariate classification and clustering
4.2.12.3 Multi-band image classification
4.2.12.4 Uncertainty and image processing
4.2.12.5 Hyperspectral image classification
4.2.13 Boundaries and zone membership
4.2.13.3 Minimum Bounding Rectangles (MBRs)
4.2.13.5 Breaklines and natural boundaries
4.2.14 Tessellations and triangulations
4.2.14.1 Delaunay Triangulation
4.2.14.2 TINs — Triangulated irregular networks
4.2.14.3 Voronoi/Thiessen polygons
4.3 Queries, Computations and Density
4.3.1 Spatial selection and spatial queries
4.3.3 Ratios, indices, normalization, standardization and rate smoothing
4.3.4 Density, kernels and occupancy
4.3.4.2 Kernel density for networks
4.3.4.3 Line and intersection densities
4.4.1.3 Extended Euclidean and Lp-metric distances
4.4.2.1 Accumulated cost surfaces and least cost paths
4.5.1 Directional analysis — overview
4.5.2 Directional analysis of linear datasets
4.5.3 Directional analysis of point datasets
4.5.4 Directional analysis of surfaces
4.6 Grid Operations and Map Algebra
4.6.1 Operations on single and multiple grids
4.6.2 Linear spatial filtering
4.6.3 Non-linear spatial filtering
5 Data Exploration and Spatial Statistics
5.1 Statistical Methods and Spatial Data
5.2 Exploratory Spatial Data Analysis
5.2.3 Cross tabulations and conditional choropleth plots
5.2.4 ESDA and mapped point data
5.2.5 Trend analysis of continuous data
5.2.6 Cluster hunting and scan statistics
5.3.1 Overview of grid-based statistics
5.3.2 Crosstabulated grid data, the Kappa Index and Cramer’s V statistic
5.3.3 Quadrat analysis of grid datasets
5.3.4.1 Non-spatial landscape metrics
5.3.4.2 Spatial landscape metrics
5.4 Point Sets and Distance Statistics
5.4.1 Basic distance-derived statistics
5.4.2 Nearest neighbor methods
5.4.4 Hot spot and cluster analysis
5.4.4.1 Hierarchical nearest neighbor clustering
5.4.4.3 Kernel density clustering
5.4.4.4 Spatio-temporal clustering
5.4.5 Proximity matrix comparisons
5.5.1 Autocorrelation, time series and spatial analysis
5.5.2 Global spatial autocorrelation
5.5.2.1 Join counts and the analysis of nominal-valued spatial data
5.5.2.3 Weighting models and lags
5.5.3 Local indicators of spatial association (LISA)
5.5.4 Significance tests for autocorrelation indices
5.6.2 Simple regression and trend surface modeling
5.6.3 Geographically Weighted Regression (GWR)
5.6.4 Spatial autoregressive and Bayesian modeling
5.6.4.1 Spatial autoregressive modeling
5.6.4.2 Conditional autoregressive and Bayesian modeling
5.6.5 Spatial filtering models
6.1.6 Statistical and fractal models
6.2.1 Gradient, slope and aspect
6.2.2.1 Profiles and cross-sections
6.2.2.2 Curvature and morphometric analysis
6.2.2.6 Longitudinal and cross-sectional curvature
6.2.2.7 Mean, maximum and minimum curvature
6.3.1 Viewsheds and RF propagation
6.3.3 Isovist analysis and space syntax
6.4.1 Overview of watersheds and drainage
6.4.4 Drainage modeling case study
6.4.4.2 Stream network construction
6.4.4.3 Stream basin construction
6.5 Gridding, Interpolation and Contouring
6.5.1 Overview of gridding and interpolation
6.5.2 Gridding and interpolation methods
6.6 Deterministic Interpolation Methods
6.6.1 Inverse distance weighting (IDW)
6.6.4 Radial basis and spline functions
6.6.6 Triangulation with linear interpolation
6.6.7 Triangulation with spline-like interpolation
6.6.8 Rectangular or bi-linear interpolation
6.6.14 Topogrid/Topo to raster
6.7 Geostatistical Interpolation Methods
6.7.1.2 Geostatistical references
6.7.1.9 Sill, range and nugget
6.7.1.12 Indicator semivariance
6.7.1.14 Comments on geostatistical software packages
6.7.1.15 Semivariance modeling
6.7.1.17 Madograms and Rodograms
6.7.1.18 Periodograms and Fourier analysis
6.7.2.6 Median-Polishing and Kriging
6.7.2.10 Non-stationary Modeling and Stratified Kriging
6.7.2.13 Conditional simulation
7 Network and Location Analysis
7.1 Introduction to Network and Location Analysis
7.1.1 Overview of network and location analysis
7.1.4 Algorithms and computational complexity theory
7.2 Key Problems in Network and Location Analysis
7.2.1 Overview — network analysis
7.2.1.1 Key problems in network analysis
7.2.1.2 Network analysis software
7.2.1.3 Key problems in location analysis
7.2.2 Heuristic and meta-heuristic algorithms
7.2.2.1 Greedy heuristics and local search
7.2.2.2 Interchange heuristics
7.2.2.5 Cross-entropy (CE) methods
7.2.2.7 Lagrangian multipliers and Lagrangian relaxation
7.2.2.8 Ant systems and ant colony optimization (ACO)
7.3 Network Construction, Optimal Routes and Optimal Tours
7.3.4 Shortest (network) path problems
7.3.4.1 Overview of shortest path problems
7.3.4.5 GIS implementations of SPAs
7.3.4.6 Further SPAs applications
7.3.5 Tours, travelling salesman problems and vehicle routing
7.3.5.1 Capacitated vehicle routing
7.4 Location and Service Area Problems
7.4.2 Larger p-median and p-center problems
7.4.2.3 Comparison of alternative p-median heuristics
7.5.1 Network traversal problems
8 Geocomputational methods and modeling
8.1 Introduction to Geocomputation
8.1.1 Geocomputational methods
8.1.2 Modeling dynamic processes within GIS
8.1.2.1 Representing time and change within GIS
8.1.2.2 Linkage/coupling versus integration/embedding
8.2.1 Introduction to geosimulation
8.2.3 Agents and agent-based models
8.2.4 Applications of agent-based models
8.2.5 Advantages of agent-based models
8.2.6 Limitations of agent-based models
8.2.7 Explanation or prediction?
8.2.8 Developing an agent-based model
8.2.9 Types of simulation/modeling (s/m) systems for agent-based modeling
8.2.10 Guidelines for choosing a simulation/modeling (s/m) system
8.2.11 Simulation/modeling (s/m) systems for agent-based modeling
8.2.12 Verification and calibration of agent-based models
8.2.13 Validation and analysis of agent-based model outputs
8.3 Artificial Neural Networks (ANN)
8.3.1 Introduction to artificial neural networks
8.3.1.1 Multi-level perceptrons (MLP)
8.3.1.2 Learning and back-propagation for MLPs
8.3.1.3 MLP Example 1: Function approximation
8.3.1.4 MLP Example 2: Landcover change modeling (LCM)
8.3.1.5 MLP Example 3: Spatial interaction modeling
8.3.2 Radial basis function networks
8.3.3 Self organizing networks
8.3.3.1 Self Organizing Maps (SOMs)
8.3.3.2 SOM unsupervised classification of hyper-spectral image data
8.3.3.3 TSP optimization using SOM concepts
8.4 Genetic Algorithms and Evolutionary Computing
8.4.1 Genetic algorithms — introduction
8.4.2 Genetic algorithm components
8.4.2.1 Encoding or representation
8.4.2.3 Population initialization
8.4.3.2 GA Example 2: Clustering
8.4.3.3 GA Example 3: Map labeling
8.4.3.4 GA Example 4: Optimum location
8.4.4 Evolutionary computing and genetic programming
R-Project spatial statistics software packages
Associations and academic bodies
Online technical dictionaries/definitions
Spatial data, test data and spatial information sources
Selected national and international data and information sources
Network analysis test datasets
Statistics and Spatial Statistics links
Figure 1‑1 3D Physical GIS models
Figure 2‑1 Attribute tables – spatial datasets
Figure 2‑2 Cyclic attribute data — Wind direction, single location
Figure 2‑3 An example map showing points, lines, and areas appropriately symbolized
Figure 2‑6 Filled contour view of field data
Figure 2‑7 Topological relationships
Figure 2‑8 Spatial weights computation
Figure 2‑9 Three alternative ways of defining neighborhood, using simple GIS functions
Figure 2‑10 Simple interpolation modeling
Figure 2‑11 Four distinct patterns of twelve points in a study area
Figure 2‑12 The process of statistical inference
Figure 2‑13 Pupil performance and a school catchment area in the East Riding of Yorkshire, UK
Figure 3‑1 Analytical process — Mitchell
Figure 3‑2 Analytical process — Draper
Figure 3‑3 PPDAC as an iterative process
Figure 3‑4 Noise map, Augsburg
Figure 3‑5 Simple GIS graphical model (ESRI ArcGIS)
Figure 3‑6 Dynamic residential growth model (Idrisi)
Figure 3‑7 Modeling wildfire risks, Arizona, USA
Figure 4‑1 Area calculation using Simpson’s rule
Figure 4‑2 3x3 grid neighborhood
Figure 4‑3 5x5 grid neighborhood
Figure 4‑4 Planimetric and surface area of a 3D triangle
Figure 4‑6 Surface model of DEM
Figure 4‑7 Smoothing techniques
Figure 4‑9 Polygon centroid (M2) and alternative polygon centers
Figure 4‑10 Center and centroid positioning
Figure 4‑11 Polygon center selection
Figure 4‑13 Point in polygon — tests and special cases
Figure 4‑14 Skeletonised convex polygon
Figure 4‑15 GRASS overlay operations, v.overlay
Figure 4‑16 Areal interpolation from census areas to a single grid cell
Figure 4‑17 Proportionally assigned population values
Figure 4‑18 Grouping data — Zone arrangement effects on voting results
Figure 4‑19 Creating postcode polygons
Figure 4‑20 Automated Zone Procedure (AZP)
Figure 4‑21 AZP applied to part of Manchester, UK
Figure 4‑22 Jenks Natural Breaks algorithm
Figure 4‑23 SPOT Band 1 image histogram — distinct peaks highlighted (CLUSTER)
Figure 4‑24 2D map of Cuprite mining district, Western Nevada, USA
Figure 4‑25 3D hypercube visualization of Cuprite mining district, Western Nevada, USA
Figure 4‑26 Single class assignment from spectral angle analysis
Figure 4‑27 Convex hull of sample point set
Figure 4‑29 Interpolation within “centroid” MBR
Figure 4‑30 Point locations inside and outside bounding polygon
Figure 4‑31 Sigmoidal fuzzy membership functions
Figure 4‑32 Delaunay triangulation of spot height locations
Figure 4‑33 Voronoi regions generated in ArcGIS and MATLab
Figure 4‑34 Voronoi cells for a homogeneous grid using a 3x3 distance transform
Figure 4‑35 Network-based Voronoi regions — Shibuya district, Tokyo
Figure 4‑36 Cell-by-cell or Local operations
Figure 4‑37 Map algebra: Index creation
Figure 4‑38 Normalization within ArcGIS
Figure 4‑39 Quantile map of normalized SIDS data
Figure 4‑40 Excess risk rate map for SIDS data
Figure 4‑42 Simple linear (box or uniform) kernel smoothing
Figure 4‑43 Univariate Normal kernel smoothing and cumulative densities
Figure 4‑44 Alternative univariate kernel density functions
Figure 4‑46 Kernel density map, Lung Case data, 3D visualization
Figure 4‑47 Univariate kernel density functions, unit bandwidth
Figure 4‑48 Cartogram creation using basic Dorling algorithm
Figure 4‑49 Cartogram creation using Dougenik, Chrisman and Niemeyer algorithm
Figure 4‑50 World Population as a Cartogram
Figure 4‑51 Cartograms of births data, 1974
Figure 4‑52 Hexagonal cartogram showing UK mortality data, age group 20-24
Figure 4‑53 Alternative measures of terrain distance
Figure 4‑54 Glasgow's Clockwork Orange Underground
Figure 4‑55 Great circle and constant bearing paths, Boston to Bristol
Figure 4‑57 Cost distance model
Figure 4‑58 Cost surface as grid
Figure 4‑59 Grid resolution and cost distance
Figure 4‑60 Accumulated cost surface and least cost paths
Figure 4‑61 Alternative route selection by ACS
Figure 4‑62 Steepest path vs tracked path
Figure 4‑63 3x3 Distance transformation – scan elements
Figure 4‑65 Distance transform, single point
Figure 4‑66 Urban traffic modeling
Figure 4‑67 Notting Hill Carnival routes
Figure 4‑68 Alternative routes selected by gradient constrained DT
Figure 4‑69 Hellisheiği power plant pipeline route selection
Figure 4‑70 Shortest and least time paths
Figure 4‑72 Manifold: Buffer operations
Figure 4‑73 Manifold: Buffering options
Figure 4‑74 Inverse distance decay models, α/dβ
Figure 4‑75 Exponential distance decay models, αe‑βd
Figure 4‑76 Directional analysis of streams
Figure 4‑77 Two-variable wind rose
Figure 4‑78 Standard distance circle and ellipses
Figure 4‑79 Correlated Random Walk simulation
Figure 4‑80 Slope and aspect plot, Mt St Helens data, USA
Figure 4‑81 Wind flow grid simulation using WindNinja
Figure 4‑82 Dilation and erosion operations
Figure 5‑1 Point-based sampling schemes
Figure 5‑2 Grid generation examples
Figure 5‑3 Grid sampling examples within hexagonal grid, 1 hectare area
Figure 5‑4 Random point generation examples — ArcGIS
Figure 5‑5 Random point samples on a network
Figure 5‑6 Brushing and linking, GeoDa
Figure 5‑7 Parallel coordinate plot
Figure 5‑11 Mapped box plot, GeoDa
Figure 5‑12 Conditional Choropleth mapping
Figure 5‑13 Exploratory analysis of radioactivity data
Figure 5‑14 Trend analysis of radioactivity dataset
Figure 5‑16 Texture analysis — variability
Figure 5‑17 Nearest Neighbor distribution
Figure 5‑18 Ripley’s K function computation
Figure 5‑19 Ripley K function, shown as transformed L function plot
Figure 5‑20 Thomas Poisson Cluster Process (20 clusters, SD=0.03, mean=5)
Figure 5‑21 Lung cancer incidence data
Figure 5‑22 Lung cancer NNh clusters
Figure 5‑23 KDE cancer incidence mapping
Figure 5‑24 Time series of stock price and volume data
Figure 5‑25 Join count patterns
Figure 5‑26 Join count computation
Figure 5‑27 Homogeneous and non-homogenous probability images
Figure 5‑28 Grouping and size effects
Figure 5‑29 Irregular lattice dataset
Figure 5‑30 Adjacency matrix, W
Figure 5‑31 Moran's I computation
Figure 5‑32 Revised source data
Figure 5‑33 Sample dataset and Moran I analysis
Figure 5‑34 Moran I (co)variance cloud, lag 1
Figure 5‑35 Local Moran I computation
Figure 5‑37 Significance tests for revised sample dataset
Figure 5‑38 Georgia educational attainment: GWR residuals map, Gaussian adaptive kernel
Figure 6‑1 East Sussex test surface, OS TQ81NE
Figure 6‑2 Pentland Hills test surface
Figure 6‑3 Linear regression surface fit to NT04 spot heights
Figure 6‑4 Raster file neighborhoods
Figure 6‑5 Vector models of TQ81NE
Figure 6‑6 First, second and third order mathematical surfaces
Figure 6‑7 Random and fractal grids
Figure 6‑8 Pseudo-random surfaces
Figure 6‑9 8-triangle slope computation
Figure 6‑10 Gradient and sampling resolution
Figure 6‑11 Slope computation output
Figure 6‑12 Frequency distribution of aspect values
Figure 6‑13 Aspect computation output
Figure 6‑14 Profile of NS transect, TQ81NE
Figure 6‑15 Multiple profile computation
Figure 6‑16 Surface morphology
Figure 6‑17 Path smoothing — vertical profile
Figure 6‑19 Viewshed computation
Figure 6‑20 3D Urban radio wave propagation modeling using Cellular Expert and ArcGIS
Figure 6‑21 Radio frequency viewshed
Figure 6‑22 Line of sight analysis
Figure 6‑23 Viewsheds and lines of sight on a synthetic (Gaussian) surface
Figure 6‑24 Isovist analysis, Street network, central London
Figure 6‑25 Axial lines and connectivity
Figure 6‑26 Depthmap — Gallery space visibility map
Figure 6‑27 D-Infinity flow assignment
Figure 6‑28 Flow direction and accumulation
Figure 6‑29 Stream identification
Figure 6‑30 Watersheds and basins
Figure 6‑31 Contour plots for alternative interpolation methods — generated with Surfer 8
Figure 6‑32 Linear interpolation of contours
Figure 6‑33 Contour computation output
Figure 6‑34 IDW as surface plot
Figure 6‑35 Contour plots for alternative IDW methods, OS NT04
Figure 6‑36 Natural Neighbor interpolation ― computation of weights
Figure 6‑37 Clough-Tocher TIN interpolation
Figure 6‑38 Regression fitting to test dataset OS NT04
Figure 6‑40 Sill, range and nugget
Figure 6‑41 Data transformation for Normality
Figure 6‑42 Anisotropy 2D map, zinc data
Figure 6‑43 Indicator variograms
Figure 6‑44 Variogram models — graphs
Figure 6‑45 Fractal analysis of TQ81NE
Figure 6‑46 Ordinary Kriging of zinc dataset
Figure 6‑47 Conditional simulation of untransformed zinc test dataset
Figure 7‑2 Visualization of lane/movement simulation (Dynameq)
Figure 7‑4 Minimum Spanning Tree
Figure 7‑5 Gabriel network construction
Figure 7‑6 Relative neighborhood network and related constructions
Figure 7‑7 Steiner MST construction
Figure 7‑8 Dantzig shortest path algorithm
Figure 7‑9 Salt Lake City — Sample networking problems and solutions
Figure 7‑10 Shortest obstacle-avoiding path
Figure 7‑11 MST, TSP and related problems
Figure 7‑12 Heuristic solution and dual circuit TSP examples
Figure 7‑13 Tanker delivery tours
Figure 7‑14 Optimum facility location on a network — LOLA solution
Figure 7‑15 Comparison of heuristic p-median solutions, Tripolis, Greece
Figure 7‑16 Facility location in Tripolis, Greece, planar model
Figure 7‑17 Service area definition
Figure 7‑18 Travel-time or drive-time zones
Figure 7‑19 Routing directions
Figure 8‑3 Moore and von Neumann neighborhoods
Figure 8‑4 Schelling segregation model
Figure 8‑5 Pedestrian movement simulation — Subway hall model
Figure 8‑6 Model development balance
Figure 8‑7 Geometric and Locational Features of the Notting Hill Carnival Swarm model
Figure 8‑8 RepastS point and click modeling and runtime environments
Figure 8‑9 RepastCity — importing GIS network data into Repast Simphony
Figure 8‑10 Repast Simphony — agent-based model visualized using NASA’s Worldwind
Figure 8‑11 StarLogo TNG drag and drop programming interface and 3D view of a simulation
Figure 8‑12 Outputs from OBEUS: Schelling residential dynamics model
Figure 8‑13 AgentSheets: The Boulder Mountain Biking Advisor
Figure 8‑14 An urban and transport dynamics model developed in AnyLogic (2006)
Figure 8‑15 Simple 3-5-2 feedforward artificial neural network
Figure 8‑16 MLP 3-5-2 with bias nodes
Figure 8‑17 ANN hidden node structure
Figure 8‑18 Sample activation functions
Figure 8‑19 MLP: Test data and fitted model
Figure 8‑21 Land cover, 1986, Chiquitania
Figure 8‑22 Distance raster (meters), anthropogenic disturbance, Chiquitania
Figure 8‑23 MLP Classifier — Idrisi
Figure 8‑24 Transition potential map
Figure 8‑25 MLP trip distribution model 1
Figure 8‑26 MLP trip distribution model 2
Figure 8‑27 Radial basis function NN model
Figure 8‑29 SOM classification of remotely-sensed hyperspectral data
Figure 8‑30 Self Organizing Map (SOM) classification — Idrisi
Figure 8‑31 SOM classified 3-band image
Figure 8‑32 Rank score transform
List of Tables
Table 1‑1 Selected terminology
Table 1‑2 Notation and symbology
Table 1‑3 Common formulas and statistical measures
Table 4‑1 Geographic data models
Table 4‑2 OGC OpenGIS Simple Features Specification — Principal Methods
Table 4‑3 Spatial overlay methods, Manifold GIS
Table 4‑4 Regional employment data — grouping affects
Table 4‑5 Selected univariate classification schemes
Table 4‑6 Image classification facilities — Selected classifiers
Table 4‑7 Selected MATLab/GRASS planar geometric analysis functions
Table 4‑8 Widely used univariate kernel density functions
Table 4‑9 Interpretation of p-values
Table 4‑10 3x3 Chamfer metrics
Table 4‑11 Linear spatial filters
Table 5‑1 Implications of Data Models
Table 5‑2 Description of methods for analysis of spatial data in ecology
Table 5‑4 Sample statistical tools for grid data — Idrisi
Table 5‑5 Simple 2-way contingency table
Table 5‑6 Simple Chi-square frequency table computation
Table 5‑7 NN Statistics and study area size
Table 5‑8 Join count analysis results
Table 5‑9 Join count mean and variance formulas
Table 5‑10 Tabulated lattice data
Table 5‑11 Selected regression analysis terminology
Table 5‑12 Georgia dataset — global regression estimates and diagnostics
Table 5‑13 Georgia dataset — comparative regression estimates and diagnostics
Table 6‑1 Morphometric features — a simplified classification
Table 6‑2 Gridding and interpolation methods
Table 6‑3 Variogram models (univariate, isotropic)
Table 7‑1 Network analysis terminology
Table 7‑2 Some key optimization problems in network analysis
Table 7‑3 Sample network analysis problem parameters
Table 7‑4 Routing functionality in selected logistics software packages
Table 7‑5 Taxonomy of location analysis problems
Table 8‑1 Agent-based modeling and GIS coupling
Table 8‑2 Agents and environments
Table 8‑3 Comparison of open source simulation/modeling toolkits
Table 8‑4 Comparison of shareware/freeware simulation/modeling systems
Table 8‑5 Comparison of proprietary simulation/modeling systems
Table 8‑6 W weights matrix, Chiquitania MLP model
Table 8‑7 SOM neighborhood and learning rate functions
|
|