Statistics and Computing

, Volume 25, Issue 5, pp 877–892 | Cite as

Comparing composite likelihood methods based on pairs for spatial Gaussian random fields



In the last years there has been a growing interest in proposing methods for estimating covariance functions for geostatistical data. Among these, maximum likelihood estimators have nice features when we deal with a Gaussian model. However maximum likelihood becomes impractical when the number of observations is very large. In this work we review some solutions and we contrast them in terms of loss of statistical efficiency and computational burden. Specifically we focus on three types of weighted composite likelihood functions based on pairs and we compare them with the method of covariance tapering. Asymptotic properties of the three estimation methods are derived. We illustrate the effectiveness of the methods through theoretical examples, simulation experiments and by analyzing a data set on yearly total precipitation anomalies at weather stations in the United States.


Covariance estimation Geostatistics Large datasets Tapering 


  1. Aho, A.V., Hopcroft, J.E., Ullman, J.D.: The Design and Analysis of Computer Algorithms. Addison-Wesley, Reading, MA (1974)MATHGoogle Scholar
  2. Banerjee, S., Gelfand, A.E., Finley, A.O., Sang, H.: Gaussian predictive process models for large spatial data sets. J. R. Stat. Soc. Ser. B 70, 825–848 (2008)MathSciNetCrossRefMATHGoogle Scholar
  3. Bevilacqua, M., Gaetan, C., Mateu, J., Porcu, E.: Estimating space and space–time covariance functions for large data sets: a weighted composite likelihood approach. J. Am. Stat. Assoc. 107, 268–280 (2012)MathSciNetCrossRefMATHGoogle Scholar
  4. Caragea, P., Smith, R.: Approximate likelihoods for spatial processes. Department of Statistics, Iowa State University, Technical report (2006)Google Scholar
  5. Cressie N.: Statistics for Spatial Data, revised edn. Wiley, New York (1993)Google Scholar
  6. Cressie, N., Johannesson, G.: Fixed rank kriging for very large spatial data sets. J. R. Stat. Soc. Ser. B 70, 209–226 (2008)MathSciNetCrossRefMATHGoogle Scholar
  7. Curriero, F., Lele, S.: A composite likelihood approach to semivariogram estimation. J. Agric. Biol. Environ. Stat. 4, 9–28 (1999)MathSciNetCrossRefGoogle Scholar
  8. Davis, R., Yau, C.-Y.: Comments on pairwise likelihood in time series models. Stat. Sin. 21, 255–277 (2011)MathSciNetMATHGoogle Scholar
  9. Doukhan, P.: Mixing Properties and Examples. Springer, New York (1994)MATHGoogle Scholar
  10. Du, J., Zhang, H., Mandrekar, V.S.: Fixed-domain asymptotic properties of tapered maximum likelihood estimators. Ann. Stat. 37, 3330–3361 (2009)MathSciNetCrossRefMATHGoogle Scholar
  11. Eidsvik, J., Reich, B., Wheeler, M., Niemi, J.: Estimation and prediction in spatial models with block composite likelihoods’. J. Comput. Gr. Stat. 15(3), 502–523 (2013)Google Scholar
  12. Fuentes, M.: Approximate likelihood for large irregularly spaced spatial data. J. Am. Stat. Assoc. 102, 321–331 (2007)Google Scholar
  13. Furrer, R., Sain, S.R.: Spam: a sparse matrix R package with emphasis on MCMC methods for Gaussian Markov random rields. J. Stat. Softw. 36, 1–25 (2010) Google Scholar
  14. Gneiting, T.: Compactly supported correlation functions. J. Multivar. Anal. 83, 493–508 (2002)MathSciNetCrossRefMATHGoogle Scholar
  15. Gneiting, T., Raftery, A.E.: Strictly proper scoring rules, prediction, and estimation. J. Am. Stat. Assoc. 102, 359–378 (2007)MathSciNetCrossRefMATHGoogle Scholar
  16. Guyon, X.: Random Fields on a Network. Springer, New York (1995)MATHGoogle Scholar
  17. Heagerty, P., Lumley, T.: Window subsampling of estimating functions with application to regression models. J. Am. Stat. Assoc. 95, 197–211 (2000)MathSciNetCrossRefMATHGoogle Scholar
  18. Huang, C., Zhang, H., Robeson, S.: On the validity of commonly used covariance and variogram functions on the sphere. Math. Geosci. 43, 721–733 (2011)MathSciNetCrossRefMATHGoogle Scholar
  19. Jenish, N., Prucha, I.R.: Central limit theorems and uniform laws of large numbers for arrays of random fields. J. Econ. 150, 86–98 (2009)MathSciNetCrossRefGoogle Scholar
  20. Joe, H., Lee, Y.: On weighting of bivariate margins in pairwise likelihood. J. Multivar. Anal. 100, 670–685 (2009)MathSciNetCrossRefMATHGoogle Scholar
  21. Kaufman, C.G., Schervish, M.J., Nychka, D.W.: Covariance tapering for likelihood-based estimation in large spatial data sets. J. Am. Stat. Assoc. 103, 1545–1555 (2008)MathSciNetCrossRefMATHGoogle Scholar
  22. Lee, Y., Lahiri, S.: Least squares variogram fitting by spatial subsampling. J. R. Stat. Soc. B 64, 837–854 (2002)MathSciNetCrossRefMATHGoogle Scholar
  23. Lindgren, F., Rue, H., Lindström, J.: An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach. J. R. Stat. Soc. Ser. B 73, 423–498 (2011)CrossRefMATHGoogle Scholar
  24. Lindsay, B.: Composite likelihood methods. Contemp. Math. 80, 221–239 (1988)MathSciNetCrossRefGoogle Scholar
  25. Lindsay, B.G., Yi, G.Y., Sun, J.: Issues and strategies in the selection of composite likelihoods. Stat. Sin. 21, 71–105 (2011)MathSciNetMATHGoogle Scholar
  26. Mardia, K.V., Marshall, J.: Maximum likelihood estimation of models for residual covariance in spatial regression. Biometrika 71, 135–146 (1984)MathSciNetCrossRefMATHGoogle Scholar
  27. Newey, W.K.: Uniform convergence in probability and stochastic equicontinuity. Econometrica 59, 1161–1167 (1991)MathSciNetCrossRefMATHGoogle Scholar
  28. Padoan, S., Bevilacqua, M.: CompRandFld: Composite-likelihood Based Analysis of Random Fields. R package version 1.0.3, (2013)Google Scholar
  29. Rue, H., Tjelmeland, H.: Fitting Gaussian Markov random fields to Gaussian fields. Scand. J. Stat. 29, 31–49 (2002)MathSciNetCrossRefMATHGoogle Scholar
  30. Shaby, B., Ruppert, D.: Tapered covariance: Bayesian estimation and asymptotics. J. Comput. Gr. Stat. 21, 433–452 (2012)MathSciNetCrossRefGoogle Scholar
  31. Stein, M.: A modeling approach for large spatial datasets. J. Korean Stat. Soc. 37, 3–10 (2008)CrossRefMATHGoogle Scholar
  32. Stein, M.: Statistical properties of covariance tapers. J. Comput. Gr. Stat. 1, 91–110 (2013)Google Scholar
  33. Stein, M., Chi, Z., Welty, L.: Approximating likelihoods for large spatial data sets. J. R. Stat. Soc. B 66, 275–296 (2004)MathSciNetCrossRefMATHGoogle Scholar
  34. Varin, C., Reid, N., Firth, D.: An overview of composite likelihood methods. Stat. Sin. 21, 5–42 (2011)MathSciNetMATHGoogle Scholar
  35. Vecchia, A.: Estimation and model identification for continuous spatial processes. J. R. Stat. Soc. B 50, 297–312 (1988)MathSciNetGoogle Scholar
  36. Wendland, H.: Piecewise polynomial, positive definite and compactly supported radial functions of minimal degree. Adv. Comput. Math. 4, 389–396 (1995)MathSciNetCrossRefMATHGoogle Scholar
  37. Whittle, P.: On stationary processes in the plane. Biometrika 49, 305–314 (1954)MathSciNetCrossRefGoogle Scholar
  38. Zhang, H., Wang, Y.: Kriging and cross-validation for massive spatial data. Environmetrics 21, 290–304 (2010)MathSciNetCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2014

Authors and Affiliations

  1. 1.DEUVUniversidad de ValparaisoValparaisoChile
  2. 2.DAISUniversità Ca’ Foscari-VeneziaVeneziaItaly

Personalised recommendations