Improving exploration of posterior distributions in spatial models - a Markov chain Monte Carlo approach

Abstract

A Markov chain Monte Carlo (MCMC) algorithm is proposed for the evaluation of a posterior distribution. The posterior distribution is from a model that has a spatial structure and exhibits many characterisics which are typically cumbersome to MCMC algorithms. The algorithm is construct with the purpose of conquering or significantly reducing these difficulties. The performance of this algorithm is then investigated for a diversity of circumstances.