Handbook of Environmental and Ecological Statistics.

Format:

Book

Author/Creator:

Gelfand, Alan E.

Contributor:

Fuentes, Montserrat.

Hoeting, Jennifer A.

Smith, Richard Lyttleton.

Language:

English

Subjects (All):

Environmental sciences--Statistical methods.

Environmental sciences.

Ecology--Statistical methods.

Ecology.

Physical Description:

1 online resource (882 pages)

Edition:

1st ed.

Place of Publication:

Milton : CRC Press LLC, 2019.

Summary:

This handbook focuses on the enormous literature applying statistical methodology and modelling to environmental and ecological processes.

Contents:

Cover

Half Title

Title Page

Copyright Page

Table of Contents

Preface

1: Introduction

I: Methodology for Statistical Analysis of Environmental Processes

2: Modeling for environmental and ecological processes

2.1 Introduction

2.2 Stochastic modeling

2.3 Basics of Bayesian inference

2.3.1 Priors

2.3.2 Posterior inference

2.3.3 Bayesian computation

2.4 Hierarchical modeling

2.4.1 Introducing uncertainty

2.4.2 Random effects and missing data

2.5 Latent variables

2.6 Mixture models

2.7 Random effects

2.8 Dynamic models

2.9 Model adequacy

2.10 Model comparison

2.10.1 Bayesian model comparison

2.10.2 Model comparison in predictive space

2.11 Summary

3: Time series methodology

3.1 Introduction

3.2 Time series processes

3.3 Stationary processes

3.3.1 Filtering preserves stationarity

3.3.2 Classes of stationary processes

3.3.2.1 IID noise and white noise

3.3.2.2 Linear processes

3.3.2.3 Autoregressive moving average processes

3.4 Statistical inference for stationary series

3.4.1 Estimating the process mean

3.4.2 Estimating the ACVF and ACF

3.4.3 Prediction and forecasting

3.4.4 Using measures of correlation for ARMA model identification

3.4.5 Parameter estimation

3.4.6 Model assessment and comparison

3.4.7 Statistical inference for the Canadian lynx series

3.5 Nonstationary time series

3.5.1 A classical decomposition for nonstationary processes

3.5.2 Stochastic representations of nonstationarity

3.6 Long memory processes

3.7 Changepoint methods

3.8 Discussion and conclusions

4: Dynamic models

4.1 Introduction

4.2 Univariate Normal Dynamic Linear Models (NDLM)

4.2.1 Forward learning: the Kalman filter

4.2.2 Backward learning: the Kalman smoother

4.2.3 Integrated likelihood.

4.2.4 Some properties of NDLMs

4.2.5 Dynamic generalized linear models (DGLM)

4.3 Multivariate Dynamic Linear Models

4.3.1 Multivariate NDLMs

4.3.2 Multivariate common-component NDLMs

4.3.3 Matrix-variate NDLMs

4.3.4 Hierarchical dynamic linear models (HDLM)

4.3.5 Spatio-temporal models

4.4 Further aspects of spatio-temporal modeling

4.4.1 Process convolution based approaches

4.4.2 Models based on stochastic partial differential equations

4.4.3 Models based on integro-difference equations

5: Geostatistical Modeling for Environmental Processes

5.1 Introduction

5.2 Elements of point-referenced modeling

5.2.1 Spatial processes, covariance functions, stationarity and isotropy

5.2.2 Anisotropy and nonstationarity

5.2.3 Variograms

5.3 Spatial interpolation and kriging

5.4 Summary

6: Spatial and spatio-temporal point processes in ecological applications

6.1 Introduction - relevance of spatial point processes to ecology

6.2 Point processes as mathematical objects

6.3 Basic definitions

6.4 Exploratory analysis - summary characteristics

6.4.1 The Poisson process-a null model

6.4.2 Descriptive methods

6.4.3 Usage in ecology

6.5 Point process models

6.5.1 Modelling environmental heterogeneity - inhomogeneous Poisson processes and Cox processes

6.5.2 Modelling clustering - Neyman Scott processes

6.5.3 Modelling inter-individual interaction - Gibbs processes

6.5.4 Model fitting - approaches and software

6.5.4.1 Approaches

6.5.4.2 Relevant software packages

6.6 Point processes in ecological applications

6.7 Marked point processes - complex data structures

6.7.1 Different roles of marks in point patterns

6.7.2 Complex models - dependence between marks and patterns

6.7.3 Marked point pattern models reflecting the sampling process.

6.8 Modelling partially observed point patterns

6.8.1 Point patterns observed in small subareas

6.8.2 Distance sampling

6.9 Discussion

6.9.1 Spatial point processes and geo-referenced data

6.9.2 Spatial point process modeling and statistical ecology

6.9.3 Other data structures

6.9.3.1 Telemetry data

6.9.3.2 Spatio-temporal patterns

6.9.4 Conclusion

6.10 Acknowledgments

7: Data assimilation

7.1 Introduction

7.2 Algorithms for data assimilation

7.2.1 Optimal interpolation

7.2.2 Variational approaches

7.2.3 Sequential approaches: the Kalman filter

7.3 Statistical approaches to data assimilation

7.3.1 Joint modeling approaches

7.3.2 Regression-based approaches

8: Univariate and Multivariate Extremes for the Environmental Sciences

8.1 Extremes and Environmental Studies

8.2 Univariate Extremes

8.2.1 Theoretical underpinnings

8.2.2 Modeling Block Maxima

8.2.3 Threshold exceedances

8.2.4 Regression models for extremes

8.2.5 Application: Fitting a time-varying GEV model to climate model output

8.2.5.1 Analysis of individual ensembles and all data

8.2.5.2 Borrowing strength across locations

8.3 Multivariate Extremes

8.3.1 Multivariate EVDs and componentwise block maxima

8.3.2 Multivariate threshold exceedances

8.3.3 Application: Santa Ana winds and dryness

8.3.3.1 Assessing tail dependence

8.3.3.2 Risk region occurrence probability estimation

8.4 Conclusions

9: Environmental Sampling Design

9.1 Introduction

9.2 Sampling Design for Environmental Monitoring

9.2.1 Design framework

9.2.2 Model-based design

9.2.2.1 Covariance estimation-based criteria

9.2.2.2 Prediction-based criteria

9.2.2.3 Mean estimation-based criteria

9.2.2.4 Multi-objective and entropy-based criteria

9.2.3 Probability-based spatial design.

9.2.3.1 Simple random sampling

9.2.3.2 Systematic random sampling

9.2.3.3 Stratified random sampling

9.2.3.4 Variable probability sampling

9.2.4 Space-filling designs

9.2.5 Design for multivariate data and stream networks

9.2.6 Space-time designs

9.2.7 Discussion

9.3 Sampling for Estimation of Abundance

9.3.1 Distance sampling

9.3.1.1 Standard probability-based designs

9.3.1.2 Adaptive distance sampling designs

9.3.1.3 Designed distance sampling experiments

9.3.2 Capture-recapture

9.3.2.1 Standard capture-recapture

9.3.2.2 Spatial capture-recapture

9.3.3 Discussion

10: Accommodating so many zeros: univariate and multivariate data

10.1 Introduction

10.2 Basic univariate modeling ideas

10.2.1 Zeros and ones

10.2.2 Zero-inflated count data

10.2.2.1 The k-ZIG

10.2.2.2 Properties of the k-ZIG model

10.2.2.3 Incorporating the covariates

10.2.2.4 Model fitting and inference

10.2.2.5 Hurdle models

10.2.3 Zeros with continuous density G(y)

10.3 Multinomial trials

10.3.1 Ordinal categorical data

10.3.2 Nominal categorical data

10.4 Spatial and spatio-temporal versions

10.5 Multivariate models with zeros

10.5.1 Multivariate Gaussian models

10.5.2 Joint species distribution models

10.5.3 A general framework for zero-dominated multivariate data

10.5.3.1 Model elements

10.5.3.2 Specific data types

10.6 Joint Attribute Modeling Application

10.6.1 Host state and its microbiome composition

10.6.2 Forest traits

10.7 Summary and Challenges

11: Gradient Analysis of Ecological Communities (Ordination)

11.1 Introduction

11.2 History of ordination methods

11.3 Theory and background

11.3.1 Properties of community data

11.3.2 Coenospace

11.3.3 Alpha, beta, gamma diversity

11.3.4 Ecological similarity and distance.

11.4 Why ordination?

11.5 Exploratory analysis and hypothesis testing

11.6 Ordination vs. Factor Analysis

11.7 A classification of ordination

11.8 Informal techniques

11.9 Distance-based techniques

11.9.1 Polar ordination

11.9.1.1 Interpretation of ordination scatter plots

11.9.2 Principal coordinates analysis

11.9.3 Nonmetric Multidimensional Scaling

11.10 Eigenanalysis-based indirect gradient analysis

11.10.1 Principal Components Analysis

11.10.2 Correspondence Analysis

11.10.3 Detrended Correspondence Analysis

11.10.4 Contrast between DCA and NMDS

11.11 Direct gradient analysis

11.11.1 Canonical Correspondence Analysis

11.11.2 Environmental variables in CCA

11.11.3 Hypothesis testing

11.11.4 Redundancy Analysis

11.12 Extensions of direct ordination

11.13 Conclusions

II: Topics in Ecological Processes

12: Species distribution models

12.1 Aims of species distribution modelling

12.2 Example data used in this chapter

12.3 Single species distribution models

12.4 Joint species distribution models

12.4.1 Shared responses to environmental covariates

12.4.2 Statistical co-occurrence

12.5 Prior distributions

12.6 Acknowledgments

13: Capture-Recapture and distance sampling to estimate population sizes

13.1 Basic ideas

13.2 Inference for closed populations

13.2.1 Censuses and finite population sampling

13.2.2 The problem of imperfect detection

13.2.3 Capture-recapture on closed populations

13.2.4 Distance sampling methods on closed populations

13.2.5 N-mixture models for closed populations

13.2.6 Count regression

13.3 Inference for open populations

13.3.1 Crosbie-Manly-Schwarz-Arnason model

13.3.2 Cormack-Jolly-Seber model and tag-recovery models

13.3.3 Pollock's robust design.

13.3.4 Capture recapture models for population growth rate.

Notes:

Description based on print version record and CIP data provided by publisher; resource not viewed.

Description based on publisher supplied metadata and other sources.

ISBN:

1-315-15250-9

1-351-64854-3

9781315152509

OCLC:

994595046