April 15, 2014
Dr. Mikyoung Jun
Professor of Statistics, Texas A&M
Title: A class of Matern-like covariance functions for smooth processes on a sphere
Abstract:
There have been noticeable advancements in developing parametric covariance models for spatial and spatio-temporal data with various applications to environmental problems. However, literature on covariance models for processes defined on the surface of a sphere with great circle distance as a distance metric is still sparse, due to its mathematical difficulties. It is known that the popular Matern covariance function, with smoothness parameter greater than 0.5, is not valid for processes on the surface of a sphere with great circle distance. We introduce an approach to produce Matern-like covariance functions for smooth processes on a sphere that are valid with great circle distance. The resulting model is isotropic and positive definite on a sphere with great circle distance, with a natural extension for nonstationarity case. We present extensive numerical comparisons of our model, with a Matern covariance model using great circle distance as well as chordal distance. We apply our new covariance model class to sea level pressure data, known to be smooth compared to other climate variables, from the CMIP5 climate model outputs.
This is a joint work with Jaehong Jeong.