A0506
Title: Nonstationary Gaussian scale mixtures for spatial extremes
Authors: Yen-Shiu Chin - National Tsing Hua University (Taiwan) [presenting]
Nan-Jung Hsu - National Tsing Hua University (Taiwan)
Hsin-Cheng Huang - Academia Sinica (Taiwan)
Abstract: Modelling spatial extremes are considered using Gaussian scale mixture models. This type of model includes asymptotic spatial dependence and asymptotic spatial independence scenarios according to the tail decay rates for their random scales. However, existing Gaussian scale mixture models are restricted to be stationary (or even isotropic), which is inappropriate for many spatial extreme applications. A nonstationary Gaussian scale mixture model using a basis-function approach is proposed. The maximum likelihood (ML) method is employed to estimate the model parameters and the ML estimators are computed via an expectation-maximization algorithm. The performance of the model is compared with stationary Gaussian scale mixture models through several simulated examples.