Prof. Nourddine Azzaoui
(Université Blaise Pascal, France)
Title
Dependance structure analysis of some stable processes in spatio-temporal models
Abstract
Spatio-temporal modeling is a fundamental step for understanding the mechanisms that govern the evolution of natural phenomena and their underlying intrinsic processes; It allows us to predict a studied variable at unobserved points of a given field based on measurements of this variable at some points of the domain. Many results and solutions have been found for second order stochastic processes, e.g. spatial kriging and spatio-temporal interpolation based on Guassian processes. We sketch a generalisation of these techniques to the case where the Gaussian spatial stochastic processes is replaced by a dependent heavy tailed alpha-stable processes. We investigate the covariation depenence structure and show how it can be used for data analysis in this context.