A nonparametric test for spatial isotropy using subsampling

Yongtao Guan, Michael Sherman, James A. Calvin

Research output: Contribution to journalArticle

44 Citations (Scopus)

Abstract

A common requirement for spatial modeling is the development of an appropriate correlation structure. Although the assumption of isotropy is often made for this structure, it is not always appropriate. A conventional practice when checking for isotropy is to informally assess plots of direction-specific sample (semi)variograms. Although a useful diagnostic, these graphical techniques are difficult to assess and open to interpretation. Formal alternatives to graphical diagnostics are valuable, but have been applied to a limited class of models. In this article we propose a formal approach to test for isotropy that is both objective and valid for a wide class of models. This approach, which is based on the asymptotic joint normality of the sample variogram, can be used to compare sample variograms in multiple directions. An L 2-consistent subsampling estimator for the asymptotic covariance matrix of the sample variogram is derived and used to construct a test statistic. A subsampling approach and a limiting chi-squared approach are developed to obtain p values of the test. Our testing approach is purely nonparametric in that no explicit knowledge of the marginal or joint distribution of the process is needed. In addition, the shape of the random field can be quite irregular. The results apply to regularly spaced data as well as to irregularly spaced data when the point locations are generated by a homogeneous Poisson process. A data example and simulation experiments demonstrate the efficacy of the approach.

Original languageEnglish (US)
Pages (from-to)810-821
Number of pages12
JournalJournal of the American Statistical Association
Volume99
Issue number467
DOIs
StatePublished - Sep 2004

Fingerprint

Subsampling
Non-parametric test
Isotropy
Variogram
Diagnostics
Semivariogram
Asymptotic Covariance Matrix
Spatial Modeling
Chi-squared
Point Location
Correlation Structure
p-Value
Marginal Distribution
Poisson process
Joint Distribution
Normality
Random Field
Simulation Experiment
Test Statistic
Efficacy

Keywords

  • Anisotropy
  • Isotropy
  • Spatial statistics
  • Subsampling
  • Variogram

ASJC Scopus subject areas

  • Mathematics(all)
  • Statistics and Probability

Cite this

A nonparametric test for spatial isotropy using subsampling. / Guan, Yongtao; Sherman, Michael; Calvin, James A.

In: Journal of the American Statistical Association, Vol. 99, No. 467, 09.2004, p. 810-821.

Research output: Contribution to journalArticle

Guan, Yongtao ; Sherman, Michael ; Calvin, James A. / A nonparametric test for spatial isotropy using subsampling. In: Journal of the American Statistical Association. 2004 ; Vol. 99, No. 467. pp. 810-821.
@article{c827a1b1d69d43dca084b42d189f3fd8,
title = "A nonparametric test for spatial isotropy using subsampling",
abstract = "A common requirement for spatial modeling is the development of an appropriate correlation structure. Although the assumption of isotropy is often made for this structure, it is not always appropriate. A conventional practice when checking for isotropy is to informally assess plots of direction-specific sample (semi)variograms. Although a useful diagnostic, these graphical techniques are difficult to assess and open to interpretation. Formal alternatives to graphical diagnostics are valuable, but have been applied to a limited class of models. In this article we propose a formal approach to test for isotropy that is both objective and valid for a wide class of models. This approach, which is based on the asymptotic joint normality of the sample variogram, can be used to compare sample variograms in multiple directions. An L 2-consistent subsampling estimator for the asymptotic covariance matrix of the sample variogram is derived and used to construct a test statistic. A subsampling approach and a limiting chi-squared approach are developed to obtain p values of the test. Our testing approach is purely nonparametric in that no explicit knowledge of the marginal or joint distribution of the process is needed. In addition, the shape of the random field can be quite irregular. The results apply to regularly spaced data as well as to irregularly spaced data when the point locations are generated by a homogeneous Poisson process. A data example and simulation experiments demonstrate the efficacy of the approach.",
keywords = "Anisotropy, Isotropy, Spatial statistics, Subsampling, Variogram",
author = "Yongtao Guan and Michael Sherman and Calvin, {James A.}",
year = "2004",
month = "9",
doi = "10.1198/016214504000001150",
language = "English (US)",
volume = "99",
pages = "810--821",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",
number = "467",

}

TY - JOUR

T1 - A nonparametric test for spatial isotropy using subsampling

AU - Guan, Yongtao

AU - Sherman, Michael

AU - Calvin, James A.

PY - 2004/9

Y1 - 2004/9

N2 - A common requirement for spatial modeling is the development of an appropriate correlation structure. Although the assumption of isotropy is often made for this structure, it is not always appropriate. A conventional practice when checking for isotropy is to informally assess plots of direction-specific sample (semi)variograms. Although a useful diagnostic, these graphical techniques are difficult to assess and open to interpretation. Formal alternatives to graphical diagnostics are valuable, but have been applied to a limited class of models. In this article we propose a formal approach to test for isotropy that is both objective and valid for a wide class of models. This approach, which is based on the asymptotic joint normality of the sample variogram, can be used to compare sample variograms in multiple directions. An L 2-consistent subsampling estimator for the asymptotic covariance matrix of the sample variogram is derived and used to construct a test statistic. A subsampling approach and a limiting chi-squared approach are developed to obtain p values of the test. Our testing approach is purely nonparametric in that no explicit knowledge of the marginal or joint distribution of the process is needed. In addition, the shape of the random field can be quite irregular. The results apply to regularly spaced data as well as to irregularly spaced data when the point locations are generated by a homogeneous Poisson process. A data example and simulation experiments demonstrate the efficacy of the approach.

AB - A common requirement for spatial modeling is the development of an appropriate correlation structure. Although the assumption of isotropy is often made for this structure, it is not always appropriate. A conventional practice when checking for isotropy is to informally assess plots of direction-specific sample (semi)variograms. Although a useful diagnostic, these graphical techniques are difficult to assess and open to interpretation. Formal alternatives to graphical diagnostics are valuable, but have been applied to a limited class of models. In this article we propose a formal approach to test for isotropy that is both objective and valid for a wide class of models. This approach, which is based on the asymptotic joint normality of the sample variogram, can be used to compare sample variograms in multiple directions. An L 2-consistent subsampling estimator for the asymptotic covariance matrix of the sample variogram is derived and used to construct a test statistic. A subsampling approach and a limiting chi-squared approach are developed to obtain p values of the test. Our testing approach is purely nonparametric in that no explicit knowledge of the marginal or joint distribution of the process is needed. In addition, the shape of the random field can be quite irregular. The results apply to regularly spaced data as well as to irregularly spaced data when the point locations are generated by a homogeneous Poisson process. A data example and simulation experiments demonstrate the efficacy of the approach.

KW - Anisotropy

KW - Isotropy

KW - Spatial statistics

KW - Subsampling

KW - Variogram

UR - http://www.scopus.com/inward/record.url?scp=4944260487&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=4944260487&partnerID=8YFLogxK

U2 - 10.1198/016214504000001150

DO - 10.1198/016214504000001150

M3 - Article

VL - 99

SP - 810

EP - 821

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 467

ER -