Statistical Methods for Spatial Data Analysis

ISBN-13: 9781584883227 / Angielski / Twarda / 2004 / 488 str.

Carol A. Gotway

Opis

- Covers the three main types of spatial data: geostatistical data, lattice data, and point patterns in theory and application - Adopts a modeling approach that ties the material into well known areas like linear models and Monte Carlo testing - Requires no previous exposure to measure theory and includes the necessary background material from stochastic process theory Spatial statistics -- one of the fastest growing areas of statistics -- has applications across a wide range of disciplines. Statistical Methods for Spatial Data Analysis takes a model-based approach and covers the theory and applications of all three major types of spatial data: geostatistical, lattice, and point pattern. The book is rich in applications and incorporates discussions on important recent developments in the field, such as spatial models for nonstationary data and generalized linear model extensions to spatial data. The authors also incorporate the use of SAS for problem solving and make the relevant code and various data sets available for download from the publisher's Web site.

Spis treści

INTRODUCTION The Need for Spatial Analysis Types of Spatial Data Autocorrelation-Concept and Elementary Measures Autocorrelation Functions The Effects of Autocorrelation on Statistical Inference Chapter Problems SOME THEORY ON RANDOM FIELDS Stochastic Processes and Samples of Size One Stationarity, Isotropy, and Heterogeneity Spatial Continuity and Differentiability Random Fields in the Spatial Domain Random Fields in the Frequency Domain Chapter Problems MAPPED POINT PATTERNS Random, Aggregated, and Regular Patterns Binomial and Poisson Processes Testing for Complete Spatial Randomness Second-Order Properties of Point Patterns The Inhomogeneous Poisson Process Marked and Multivariate Point Patterns Point Process Models Chapter Problems SEMIVARIOGRAM AND COVARIANCE FUNCTION ANALYSIS AND ESTIMATION Introduction Semivariogram and Covariogram Covariance and Semivariogram Models Estimating the Semivariogram Parametric Modeling Nonparametric Estimation and Modeling Estimation and Inference in the Frequency Domain On the Use of Non-Euclidean Distances in Geostatistics Supplement: Bessel Functions Chapter Problems SPATIAL PREDICTION AND KRIGING Optimal Prediction in Random Fields Linear Prediction-Simple and Ordinary Kriging Linear Prediction with a Spatially Varying Mean Kriging in Practice Estimating Covariance Parameters Nonlinear Prediction Change of Support On the Popularity of the Multivariate Gaussian Distribution Chapter Problems SPATIAL REGRESSION MODELS Linear Models with Uncorrelated Errors Linear Models with Correlated Errors Generalized Linear Models Bayesian Hierarchical Models Chapter Problems SIMULATION OF RANDOM FIELDS Unconditional Simulation of Gaussian Random Fields Conditional Simulation of Gaussian Random Fields Simulated Annealing Simulating from Convolutions Simulating Point Processes Chapter Problems NON-STATIONARY COVARIANCE Types of Non-Stationarity Global Modeling Approaches Local Stationarity SPATIO-TEMPORAL PROCESSES A New Dimension Separable Covariance Functions Non-Separable Covariance Functions The Spatio-Temporal Semivariogram Spatio-Temporal Point Processes

O autorze

Oliver Schabenberger, Carol A. Gotway