Circulant Embedding of Approximate Covariances for Inference From Gaussian Data on Large Lattices

ABSTRACT Recently proposed computationally efficient Markov chain Monte Carlo (MCMC) and Monte Carlo expectation–maximization (EM) methods for estimating covariance parameters from lattice data rely on successive imputations of values on an embedding lattice that is at least two times larger in each dimension. These methods can be considered exact in some sense, but we demonstrate that using such a large number of imputed values leads to slowly converging Markov chains and EM algorithms. We propose instead the use of a discrete spectral approximation to allow for the implementation of these methods on smaller embedding lattices. While our methods are approximate, our examples indicate that the error introduced by this approximation is small compared to the Monte Carlo errors present in long Markov chains or many iterations of Monte Carlo EM algorithms. Our results are demonstrated in simulation studies, as well as in numerical studies that explore both increasing domain and fixed domain asymptotics. We compare the exact methods to our approximate methods on a large satellite dataset, and show that the approximate methods are also faster to compute, especially when the aliased spectral density is modeled directly. Supplementary materials for this article are available online.

[1]  P. Whittle ON STATIONARY PROCESSES IN THE PLANE , 1954 .

[2]  B. Matérn Spatial variation : Stochastic models and their application to some problems in forest surveys and other sampling investigations , 1960 .

[3]  X. Guyon Parameter estimation for a stationary process on a d-dimensional lattice , 1982 .

[4]  A. Yaglom Correlation Theory of Stationary and Related Random Functions I: Basic Results , 1987 .

[5]  R. Dahlhaus,et al.  Edge effects and efficient parameter estimation for stationary random fields , 1987 .

[6]  A. V. Vecchia Estimation and model identification for continuous spatial processes , 1988 .

[7]  A. Wood,et al.  Simulation of Stationary Gaussian Processes in [0, 1] d , 1994 .

[8]  C. R. Dietrich,et al.  Fast and Exact Simulation of Stationary Gaussian Processes through Circulant Embedding of the Covariance Matrix , 1997, SIAM J. Sci. Comput..

[9]  A. T. A. Wood,et al.  Simulation of stationary Gaussian vector fields , 1999, Stat. Comput..

[10]  Michael L. Stein,et al.  Interpolation of spatial data , 1999 .

[11]  Roger Woodard,et al.  Interpolation of Spatial Data: Some Theory for Kriging , 1999, Technometrics.

[12]  Michael L. Stein,et al.  Fast and Exact Simulation of Fractional Brownian Surfaces , 2002 .

[13]  Raghu Kacker,et al.  Digital Library of Mathematical Functions , 2003 .

[14]  Sw. Banerjee,et al.  Hierarchical Modeling and Analysis for Spatial Data , 2003 .

[15]  Daniel W. Lozier,et al.  NIST Digital Library of Mathematical Functions , 2003, Annals of Mathematics and Artificial Intelligence.

[16]  Zhiyi Chi,et al.  Approximating likelihoods for large spatial data sets , 2004 .

[17]  Leonhard Held,et al.  Gaussian Markov Random Fields: Theory and Applications , 2005 .

[18]  T. Gneiting,et al.  Fast and Exact Simulation of Large Gaussian Lattice Systems in ℝ2: Exploring the Limits , 2006 .

[19]  N. Reid,et al.  AN OVERVIEW OF COMPOSITE LIKELIHOOD METHODS , 2011 .

[20]  H. Rue,et al.  An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach , 2011 .

[21]  Michael L. Stein,et al.  Bayesian and Maximum Likelihood Estimation for Gaussian Processes on an Incomplete Lattice , 2014, 1402.4281.

[22]  Montserrat Fuentes,et al.  Multivariate spatial modeling of conditional dependence in microscale soil elemental composition data , 2014 .