Covariance Models for Vector Random Fields in Space and Time

• Activity Number: 596
• Type: Topic Contributed
• Date/Time: Thursday, August 2, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics and the Environment
• Author(s): Juan Du*+ and Chunsheng Ma
• Companies: Kansas State University and Wichita State University
• Keywords: Spatial statistics ; Covariance matrix function ; Vector random field ; Gaussian random field

Abstract

Multivariate data over space and time are often observed in various disciplines, such as atmospheric sciences, meteorology, engineering and agriculture. The modeling and simulation study of this type of data call on the development of the vector random fields with various properties for both theoretical study and practical use. Several families of the covariance matrix structures through an efficient approach are proposed to model vector random fields in Gaussian and non-Gaussian cases, as well as stationary and non-stationary situations. Basic theoretical properties of the proposed multivariate covariance models are explored to provide some guidance on the practical application of these models. The model fitting and applications are demonstrated using simulation study and some weather data sets.