High?dimensional multivariate geostatistics: A Bayesian matrix?normal approach

Joint modeling of spatially oriented dependent variables is commonplace in the environmental sciences, where scientists seek to estimate relationships among a set outcomes accounting for dependence these and spatial each outcome. Such now sought massive data sets with measured at very large number locations. Bayesian inference, while attractive accommodating uncertainties through hierarchical structures, can become computationally onerous because its reliance on iterative estimation algorithms. This article develops conjugate framework analyzing multivariate using analytically tractable posterior distributions that obviate We discuss differences between response itself as process latent model. illustrate computational inferential benefits models simulation studies analysis vegetation index observations numbering millions.