|
Abstract:
|
Max-stable processes constitute a broad class of models for capturing spatial dependence among extremes, a feature common to many environmental phenomena. A drawback of these models is that the likelihood for max-stable process is unavailable for more than a small number of spatial locations, precluding Bayesian inference. The proposed hierarchical Bayesian model uses positive stable random effects to model residual spatial dependence and takes the form of two popular max-stable families as limiting cases.
|
