Weighted composite likelihood-based tests for space-time separability of covariance functions

Statistics and Computing volume 20, pages283–293(2010)Cite this article

Abstract

Testing for separability of space-time covariance functions is of great interest in the analysis of space-time data. In this paper we work in a parametric framework and consider the case when the parameter identifying the case of separability of the associated space-time covariance lies on the boundary of the parametric space. This situation is frequently encountered in space-time geostatistics. It is known that classical methods such as likelihood ratio test may fail in this case.

We present two tests based on weighted composite likelihood estimates and the bootstrap method, and evaluate their performance through an extensive simulation study as well as an application to Irish wind speeds. The tests are performed with respect to a new class of covariance functions, which presents some desirable mathematical features and has margins of the Generalized Cauchy type. We also apply the test on a element of the Gneiting class, obtaining concordant results.

This is a preview of subscription content, log in to check access.

References

  1. Berg, C., Mateu, J., Porcu, E.: The Dagum family of isotropic correlation functions. Bernoulli 14, 1134–1149 (2008)

    MATH  Article  MathSciNet  Google Scholar 

  2. Besag, J.: Spatial interaction and the statistical analysis of lattice systems (with discussions). J. R. Stat. Soc. B 36, 192–236 (1974)

    MATH  MathSciNet  Google Scholar 

  3. Bevilacqua, M., Gaetan, C., Mateu, J., Porcu, E.: Estimating space and space-time covariance functions: a weighted composite likelihood approach. J. Am. Stat. Assoc. (2009, to appear)

  4. Cox, D.R.: Partial likelihood. Biometrika 62, 269–276 (1975)

    MATH  Article  MathSciNet  Google Scholar 

  5. Cressie, N., Huang, H.: Classes of nonseparable, spatiotemporal stationary covariance functions. J. Am. Stat. Assoc. 94, 1330–1340 (1999)

    MATH  Article  MathSciNet  Google Scholar 

  6. Crujeiras, R.M., Fernández-Casal, R., González-Manteiga, W.: Goodness-of-fit tests for spatial spectral density. Technical report 07-01, Departamento de Estadística e Investigación Operativa, Universidade de Santiago de Compostela (2008)

  7. Curriero, F., Lele, S.: A composite likelihood approach to semivariogram estimation. J. Agric. Biol. Environ. Stat. 4, 9–28 (1999)

    Article  MathSciNet  Google Scholar 

  8. De Luna, X., Genton, M.: Predictive spatio-temporal models for spatially sparse environmental data. Stat. Sinica 15, 547–568 (2005)

    MATH  Google Scholar 

  9. Efron, B.: The Jackknife, the Bootstrap and Other Resampling Plans. SIAM, Philadelphia (1982)

    Google Scholar 

  10. Feng, Z.Z., McCulloch, C.E.: Using bootstrap likelihood ratios in finite mixture models. J. R. Stat. Soc. Ser. B 58, 609–617 (1996)

    MATH  Google Scholar 

  11. Fuentes, M.: Interpolation of nonstationary air pollution processes: a spatial spectral approach. Stat. Model. 2, 281–298 (2002)

    MATH  Article  MathSciNet  Google Scholar 

  12. Fuentes, M.: A formal test for nonstationarity of spatial stochastic processes. J. Multivar. Anal. 96, 30–54 (2005)

    MATH  Article  MathSciNet  Google Scholar 

  13. Fuentes, M.: Testing for separability of spatial-temporal covariance functions. J. Stat. Plan. Inference 136, 447–466 (2006)

    MATH  Article  MathSciNet  Google Scholar 

  14. Gneiting, T.: Stationary covariance functions for space-time data. J. Am. Stat. Assoc. 97, 590–600 (2002)

    MATH  Article  MathSciNet  Google Scholar 

  15. Gneiting, T., Schlather, M.: Stochastic models that separate fractal dimension and the Hurst effect. Soc. Indust. Appl. Math. Rev. 46, 269–282 (2004)

    MATH  MathSciNet  Google Scholar 

  16. Gneiting, T., Genton, M.G., Guttorp, P.: Geostatistical space-time models, stationarity, separability and full symmetry. In: Finkenstadt, B., Held, L., Isham, V. (eds.) Statistical Methods for Spatio-Temporal Systems, pp. 151–175. Chapman & Hall/CRC, Boca Raton (2007)

    Google Scholar 

  17. Godambe, V.: An optimum property of regular maximum likelihood equation. Ann. Math. Stat. 31, 1208–1211 (1960)

    Article  MathSciNet  Google Scholar 

  18. Guan, Y., Sherman, M., Calvin, J.A.: A nonparametric test for spatial isotropy using subsampling. J. Am. Stat. Assoc. 99, 810–821 (2004)

    MATH  Article  MathSciNet  Google Scholar 

  19. Hartigan, J.A.: A failure of likelihood asymptotic for normal mixtures. In: LeCam, L., Olshen, R.A. (eds.) Proceedings of the Berkeley Conference in Honor of Jerzy Neyman and Jack Kiefer, vol. II, pp. 807–810. Wadsworth and Brooks, Belmont (1985)

    Google Scholar 

  20. Haslett, J., Raftery, A.: Space-time modelling with long-memory dependence: assessing Ireland’s wind power resource. Appl. Stat. 38, 1–50 (1989)

    Article  Google Scholar 

  21. Heagerty, P., Lumley, T.: Window subsampling of estimating functions with application to regression models. J. Am. Stat. Assoc. 95, 197–211 (2000)

    MATH  Article  MathSciNet  Google Scholar 

  22. Heyde, C.: Quasi-likelihood and Its Application: A General Approach to Optimal Parameter Estimation. Springer, New York (1997)

    Google Scholar 

  23. Kolesárová, A.: Limit properties of quasi-arithmetic means. Fuzzy Sets Syst. 124, 65–71 (2001)

    MATH  Article  Google Scholar 

  24. Li, B., Genton, M.G., Sherman, M.: A nonparametric assessment of properties of space-time covariance functions. J. Am. Stat. Assoc. 102, 736–744 (2007)

    MATH  Article  MathSciNet  Google Scholar 

  25. Lindsay, B.: Composite likelihood methods. Contemp. Math. 80, 221–239 (1988)

    MathSciNet  Google Scholar 

  26. Lu, H., Zimmerman, D.L.: Testing for directional symmetry in spatial dependence using the periodogram. J. Stat. Plan. Inference 129, 369–385 (2005)

    MATH  Article  MathSciNet  Google Scholar 

  27. Ma, C.: A class of stationary random fields with a simple correlation structure. J. Multivar. Anal. 94, 313–327 (2004)

    Article  Google Scholar 

  28. Mardia, K., Marshall, R.: Maximum likelihood estimation of models for residual covariance in spatial regression. Biometrika 71, 135–146 (1984)

    MATH  Article  MathSciNet  Google Scholar 

  29. Mateu, J., Juan, P., Porcu, E.: Geostatistical analysis through spectral techniques: some words of caution. Commun. Stat. Comput. Simul. 36, 1035–1051 (2007a)

    MATH  Article  MathSciNet  Google Scholar 

  30. Mateu, J., Porcu, E., Nicolis, O.: A note on decoupling of local and global behaviour for the Dagum random field. Probab. Eng. Mech. 22, 320–329 (2007b)

    Article  Google Scholar 

  31. Mitchell, M., Genton, M.G., Gumpertz, M.: Testing for separability of space-time covariances. Environmetrics 16, 819–831 (2005)

    Article  MathSciNet  Google Scholar 

  32. Mitchell, M., Genton, M.G., Gumpertz, M.: A likelihood ratio test for separability of covariances. J. Multivar. Anal. 97, 1025–1043 (2006)

    MATH  Article  MathSciNet  Google Scholar 

  33. Porcu, E., Gregori, P., Mateu, J.: Nonseparable stationary anisotropic space time covariance functions. Stoch. Environ. Res. Risk Assess. 21, 113–122 (2006)

    Article  MathSciNet  Google Scholar 

  34. Porcu, E., Mateu, J., Christakos, G.: Quasi-arithmetic means of covariance functions with potential applications to space-time data. J. Multivar. Anal. (2009, to appear). doi:10.1016/j.jmva.2009.02.013

  35. Scaccia, L., Martin, R.J.: Testing for simplification in spatial models. In: COMPSTAT 2002, pp. 581–586. Physica, Heidelberg (2002)

    Google Scholar 

  36. Scaccia, L., Martin, R.J.: Testing axial symmetry and separability of lattice processes. J. Stat. Plan. Inference 131, 19–39 (2005)

    MATH  Article  MathSciNet  Google Scholar 

  37. Stein, M.: Space-time covariance functions. J. Am. Stat. Assoc. 100, 310–321 (2005)

    MATH  Article  Google Scholar 

  38. Varin, C., Vidoni, P.: A note on composite likelihood inference and model selection. Biometrika 52, 519–528 (2005)

    Article  MathSciNet  Google Scholar 

  39. Wang, H.H., Zhang, H.: Model-based clustering for cross-sectional time series data. J. Agric. Biol. Environ. Stat. 7, 107–127 (2002)

    Article  Google Scholar 

  40. Zhang, H.: Inconsistent estimation and asymptotically equal interpolations in model-based geostatistics. J. Am. Stat. Assoc. 99, 250–261 (2004)

    MATH  Article  Google Scholar 

  41. Zhang, H., Zimmerman, D.: Towards reconciling two asymptotic frameworks in spatial statistics. Biometrika 92, 921–936 (2005)

    MATH  Article  MathSciNet  Google Scholar 

Download references

Author information

Affiliations

  1. Dipartimento di Scienze Statistiche, Università di Padova, Padova, Italy

    M. Bevilacqua

  2. Department of Mathematics, Universitat Jaume I, Castellón, Spain

    J. Mateu & E. Porcu

  3. Department of Statistics, Purdue University, Mathematical Science Building, West Lafayette, IN, 47906, USA

    H. Zhang

  4. Department of Quantitative Methods for Business and Economic Sciences, University of Milano Bicocca, Milan, Italy

    A. Zini

Authors
  1. M. Bevilacqua
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. J. Mateu
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. E. Porcu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. H. Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. A. Zini
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to J. Mateu.

Additional information

Research supported by grant MTM2007-62923 from Spanish Ministry of Education and Science.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Bevilacqua, M., Mateu, J., Porcu, E. et al. Weighted composite likelihood-based tests for space-time separability of covariance functions. Stat Comput 20, 283–293 (2010). https://doi.org/10.1007/s11222-009-9121-3

Download citation

Keywords

  • Fractal dimension
  • Full symmetry
  • Hurst effect
  • Space-time covariance functions
  • Space-time separability
  • Weighted composite likelihood