Spatio-temporal model for Plague dynamics

Abstract

Biologists have great interest in studying sylvatic plague to give a better understanding of the process, but also to prevent plague from being transmitted to humans. I have studied the processes of sylvatic plague through a unique dataset from Kazakhstan. This dataset concerns plague among rodents and includes information of rodent abundance. The spatiotemporal structure of the data permits the detailed models. There is however much missing data. The uncertainty related to the missing data have often been ignored. By using Bayesian hierarchical methods, I have taken into account the uncertainty related to missing data. In this analysis I found abundance of rodents and average temperature of April to have significant influence on the process of plague. It also seems, one are more likely to find plague at a site the next year if there is much plague present.