Space-Time Modeling and Signal Detection

Abstract

This talk contains two parts. The first part is concerned with space-time modeling. Geostatistical approaches to spatio-temporal prediction in many scientific fields require the knowledge of its spatial and temporal dependence structure. First, I will introduce some of our proposed space-time models, including both stationary and nonstationary covariance models, and those with separable and non-separable space-time covariance functions. Some applications of these models to real data will also be presented. In the second part, a new nonparametric hypothesis-testing procedure for detecting a spatial signal is given. The proposed procedure is based on controlling the false discovery rate and a concept known as generalized degrees of freedom. Finally, I will mention some ongoing research.