Abstract:
|
In this work, we propose a flexible model for multivariate spatial extremes, which is parsimonious and designed to capture different combinations of marginal and cross-extremal dependence structures within and across different spatial random fields. In particular, our model can capture all possible distinct combinations of extremal dependence structures within each individual spatial process, while allowing for flexible cross-process extremal dependence structures, for both the upper and lower tails. The model may be seen as a multi-factor copula model, and we perform Bayesian inference using a Markov chain Monte Carlo algorithm based on carefully designed block proposals with adaptive stepsize. If time allows, we will discuss an application of our model to study the extremal dependence structure of daily average air temperature and relative humidity fields in a region of south-eastern US.
|