Local Models for Spatial Analysis, Second Edition

Written in recognition of developments in spatial data analysis that focused on differences between places, the first edition of Local Models for Spatial Analysis broke new ground with its focus on local modelling methods. Reflecting the continued growth and increased interest in this area, the second edition describes a wide range of methods which account for local variations in geographical properties. What’s new in the Second Edition: Additional material on geographically-weighted statistics and local regression approaches A better overview of local models with reference to recent critical reviews about the subject area Expanded coverage of individual methods and connections between them Chapters have been restructured to clarify the distinction between global and local methods A new section in each chapter references key studies or other accounts that support the book Selected resources provided online to support learning An introduction to the methods and their underlying concepts, the book uses worked examples and case studies to demonstrate how the algorithms work their practical utility and range of application. It provides an overview of a range of different approaches that have been developed and employed within Geographical Information Science (GIScience). Starting with first principles, the author introduces users of GISystems to the principles and application of some widely used local models for the analysis of spatial data, including methods being developed and employed in geography and cognate disciplines. He discusses the relevant software packages that can aid their implementation and provides a summary list in Appendix A. Presenting examples from a variety of disciplines, the book demonstrates the importance of local models for all who make use of spatial data. Taking a problem driven approach, it provides extensive guidance on the selection and application of local models.

IntroductionRemit Of This Book Local Models and MethodsWhat Is Local? Spatial Dependence and Autocorrelation Spatial Scale Stationarity Spatial Data Models Datasets Used for Illustrative Purposes A Note on Notation OverviewLocal ModellingStandard Methods and Local VariationsApproaches to Local Adaptation Stratification or Segmentation of Spatial Data Moving Window/Kernel Methods Locally-Varying Model Parameters Transforming and Detrending Spatial Data Categorising Local Statistical Models Local Models And Methods And The Structure Of The Book OverviewGrid DataExploring Spatial Variation in Gridded Variables Global Univariate StatisticsLocal Univariate StatisticsAnalysis of Grid DataMoving Windows for Grid AnalysisWavelets Segmentation Analysis of Digital Elevation Models OverviewSpatial Patterning in Single VariablesLocal Summary Statistics Geographically Weighted StatisticsSpatial Autocorrelation: Global MeasuresSpatial Association and Categorical DataOther Issues OverviewSpatial RelationsGlobal RegressionSpatial and Local Regression Regression and Spatial DataSpatial Autoregressive ModelsMultilevel Modelling Allowing for Local Variation in Model ParametersMoving Window Regression (Mwr) Geographically Weighted Regression (Gwr) Spatially Weighted Classification Local Regression Methods: Some Pros and ConsOverview Spatial Prediction 1: Deterministic Methods, Curve Fitting, and SmoothingPoint Interpolation Global Methods Local Methods Areal Interpolation General Approaches: Overlay Local Models and Local DataLimitations: Point And Areal InterpolationOverviewSpatial Prediction 2: GeostatisticsRandom Function ModelsStationarityExploring Spatial VariationKrigingGlobally Constant Mean: Simple Kriging Locally Constant Mean Models Ordinary Kriging Cokriging Equivalence of Splines And Kriging Conditional Simulation The Change of Support Problem Other ApproachesLocal Approaches: Nonstationary Models Nonstationary Mean Nonstationary Models For Prediction Nonstationary Variogram Variograms in Texture Analysis Summary Point Patterns and Cluster DetectionPoint Patterns Visual Examination of Point Patterns Measuring Event Intensity And Distance MethodsStatistical Tests of Point PatternsGlobal MethodsMeasuring Event Intensity .Distance MethodsOther IssuesLocal MethodsMeasuring Event Intensity Locally Accounting For The Population at Risk The Local K Function Point Patterns and Detection of Clusters OverviewSummary: Local Models for Spatial AnalysisReviewIssues Software Future Developments SummaryA SoftwareReferencesIndex

Christopher D. Lloyd