Abstract
We discuss a new class of spatially varying, simultaneous autoregressive (SVSAR) models motivated by interests in flexible, non-stationary spatial modelling scalable to higher dimensions. SVSAR models are hierarchical Markov random fields extending traditional SAR models. We develop Bayesian analysis using Markov chain Monte Carlo methods of SVSAR models, with extensions to spatio-temporal contexts to address problems of data assimilation in computer models. A motivating application in atmospheric science concerns global CO emissions where prediction from computer models is assessed and refined based on high-resolution global satellite imagery data. Application to synthetic and real CO data sets demonstrates the potential of SVSAR models in flexibly representing inhomogeneous spatial processes on lattices, and their ability to improve estimation and prediction of spatial fields. The SVSAR approach is computationally attractive in even very large problems; computational efficiencies are enabled by exploiting sparsity of high-dimensional precision matrices.
Similar content being viewed by others
References
Akaike, H., Kitagawa, G. (Eds.). (1998). The practice of time series analysis. New York: Springer-Verlag.
Anselin, L. (1988). Spatial econometrics: methods and models. Dordrecht: Kluwer Academic Publishers.
Arellano, A. F. J., Kasibhatla, P. S., Goglio, L., van der Werf, G. R., Randerson, J. T. (2004). Top-down estimates of global CO sources using MOPITT measurements. Geophysical Research Letters, 31, L0104.
Arellano, A. F., Kasibhatla, P. S., Giglio, L., van der Werf, G. R., Randerson, J. T., Collatz, G. J. (2006). Time-dependent inversion estimates of global biomass-burning CO emissions using measurement of pollution in the troposphere (MOPITT) measurements. Journal of Geophysical Research, 111, D09303.
Besag, J. (1974). Spatial interaction and the statistical analysis of lattice systems. Journal of the Royal Statistical Society (Series B, Methodological), 36, 192–236.
Besag, J., Kooperberg, C. (1995). On conditional and intrinsic autoregressions. Biometrika, 82, 733–746.
Brewer, M. J., Nolan, A. J. (2007). Variable smoothing in Bayesian intrinsic autoregressions. Environmetrics, 18, 841–857.
Chevallier, F. (2007). Impact of correlated observation errors on inverted CO\(_2\) surface fluxes from OCO measurements. Geophysical Research Letters, 34, L24804.
Chevallier, F., Engelen, R. J., Peylin, P. (2005a). The contribution of AIRS data to the estimation of CO\(_2\) sources and sinks. Geophysical Research Letters, 32, L23801.
Chevallier, F., Fisher, M., Peylin, P., Serrar, S., Bousquet, P., Breon, F. M., et al. (2005b). Inferring CO\(_2\) sources and sinks from satellite observations: Method and application to TOVS data. Journal of Geophysical Research, 110, D24309.
Chevallier, F., Breon, F. M., Rayner, P. J. (2007). Contribution of the Orbiting Carbon Observatory to the estimation of CO\(_2\) sources and sinks: Theoretical study in a variational data assimilation framework. Journal of Geophysical Research, 112, D09307.
Chevallier, F., Engelen, R. J., Carouge, C., Conway, T. J., Peylin, P., Pickett-Heaps, C., et al. (2009a). AIRS-based versus flask-based estimation of carbon surface fluxes. Journal of Geophysical Research, 114, D20303.
Chevallier, F., Maksyutov, S., Bousquet, P., Breon, F. M., Saito, R., Yoshida, Y., et al. (2009b). On the accuracy of the CO\(_2\) surface fluxes to be estimated from the GOSAT observations. Geophysical Research Letters, 36, L19807.
Dobra, A., Lenkoski, A., Rodriguez, A. (2011). Bayesian inference for general Gaussian graphical models with application to multivariate lattice data. Journal of the American Statistical Association, 106, 1418–1433.
Fuentes, M. (2002). Spectral methods for nonstationary spatial processes. Biometrika, 89, 197–210.
Gneiting, T. (2002). Compactly supported correlation functions. Journal of Multivariate Analysis, 83, 493–508.
Golub, G. H., Van Loan, C. F. (1996). Matrix computations. Baltimore: Johns Hopkins University Press.
Higdon, D. M. (1998). A process-convolution approach to modelling temperatures in the North Atlantic Ocean. Environmental and Ecological Statistics, 5, 173–190.
Higdon, D. M., Swall, J., Kern, J. (1999). Non-stationary spatial modeling. In J. M. Bernardo, J. O. Berger, A. P. Dawid, A. F. M. Smith (Eds.), Bayesian statistics (Vol. 6, pp. 761–768). Oxford: Oxford University Press.
Jones, B., West, M. (2005). Covariance decomposition in undirected graphical models. Biometrika, 92, 779–786.
Jones, B., Carvalho, C. M., Dobra, A., Hans, C., Carter, C., West, M. (2005). Experiments in stochastic computation for high-dimensional graphical models. Statistical Science, 20, 388–400.
Kitagawa, G., Gersch, W. (1996). Smoothness priors analysis of time series, Lecture Notes in Statistics (Vol. 116). New York: Springer-Verlag.
Kopacz, M., Jacob, D. J., Henze, D. K., Heald, C. L., Streets, D. G., Zhang, Q. (2009). Comparison of adjoint and analytical Bayesian inversion methods for constraining Asian sources of carbon monoxide using satellite (MOPITT) measurements of CO columns. Journal of Geophysical Research, 114, D04305.
Kopacz, M., Jacob, D. J., Fisher, J. A., Logan, J. A., Zhang, L., Megretskaia, I. A., et al. (2010). Global estimates of CO sources with high resolution by adjoint inversion of multiple satellite datasets (MOPITT, AIRS, SCIAMACHY, TES). Atmospheric Chemistry & Physics, 10, 855–876.
Lopes, H. F., Salazar, E., Gamerman, D. (2008). Spatial dynamic factor analysis. Bayesian Analysis, 3, 1–34.
Meirink, J. F., Bergamaschi, P., Frankenberg, C., d’Amelio, M. T. S., Dlugokencky, E. J., Gatti, L. V., et al. (2008). Four-dimensional variational data assimilation for inverse modeling of atmospheric methane emissions: Analysis of SCIAMACHY observations. Journal of Geophysical Research, 113, D17301.
Mukherjee, C., Kasibhatla, P. S., West, M. (2011). Bayesian statistical modeling of spatially correlated error structure in atmospheric tracer inverse analysis. Atmospheric Chemistry & Physics, 11, 5365–5382.
Prado, R., West, M. (2010). Time series: Modeling, computation and inference. London: Chapman & Hall/CRC Press, The Taylor Francis Group.
Reich, B. J., Hodges, J. S. (2008). Modeling longitudinal spatial periodontal data: A spatially adaptive model with tools for specifying priors and checking fit. Biometrics, 64, 790–799.
Rue, H. (2001). Fast sampling of Gaussian Markov random fields. Journal of the Royal Statistical Society (Series B, Methodological), 63, 325–338.
Rue, H., Held, L. (2005). Gaussian Markov random fields: Theory and applications, Monographs on Statistics and Applied Probability (Vol. 104). London: Chapman & Hall.
Sampson, P. D., Guttorp, P. (1992). Nonparametric estimation of nonstationary spatial covariance structure. Journal of the American Statistical Association, 87, 108–119.
Spiegelhalter, D., Best, N., Carlin, B. P., van der Linde, A. (2002). Bayesian measures of model complexity and fit (with discussion). Journal of the Royal Statistical Society (Series B, Methodological), 64, 583–639.
Suchard, M. A., Holmes, C., West, M. (2010a). Some of the What?, Why?, How?, Who? and Where? of graphics processing unit computing for Bayesian analysis. Bulletin of the International Society for Bayesian, Analysis, 17, 12–16.
Suchard, M. A., Wang, Q., Chan, C., Frelinger, J., Cron, A. J., West, M. (2010b). Understanding GPU programming for statistical computation: Studies in massively parallel massive mixtures. Journal of Computational and Graphical Statistics, 19, 419–438.
West, M., Harrison, P. J. (1997). Bayesian forecasting and dynamic models (2nd ed.). New York: Springer.
Whittle, P. (1954). On stationary processes in the plane. Biometrika, 41, 434–449.
Acknowledgments
We are grateful to the Editor and two anonymous referees for their positive and constructive comments on an earlier version of the paper, and to Abel Rodriguez for useful comments and suggestions. This work was partly supported by Grants from the US National Science Foundation [M.W., #DMS-1106516] and the US National Aeronautics and Space Administration [P.S.K., sub-award 2011-2654 of Grant #NNX11AF96G to the University of California at Irvine]. Any opinions, findings and conclusions or recommendations expressed in this work are those of the authors and do not necessarily reflect the views of the NSF or NASA.
Author information
Authors and Affiliations
Corresponding author
Additional information
This work was completed while C. Mukherjee was a PhD student in Statistical Science at Duke University.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Supplementary material 1 (avi 104802 KB)
Supplementary material 2 (avi 100308 KB)
About this article
Cite this article
Mukherjee, C., Kasibhatla, P.S. & West, M. Spatially varying SAR models and Bayesian inference for high-resolution lattice data. Ann Inst Stat Math 66, 473–494 (2014). https://doi.org/10.1007/s10463-013-0426-9
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10463-013-0426-9