Parallel Algorithms For Bayesian Inference In Spatial Gaussian Models

Markov chain Monte Carlo (MCMC) implementations of Bayesian inference for latent spatial Gaussian models are very computationally intensive, and restrictions on storage and computation time are limiting their application to large problems. Here we propose three parallel algorithms for linear algebra calculations in these implementations. The algorithms' performance is discussed with respect to a simulation study, which demonstrates the increase in speed and feasible problem size as a function of the number of processors. We discuss how parallel algorithms may be useful more generally in MCMC schemes for these models.