Nonparametric Cross Covariogram for Multivariate Geostatistics

• Activity Number: 118
• Type: Topic Contributed
• Date/Time: Monday, August 2, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics and the Environment
• Author(s): Hao Zhang*+
• Companies: Purdue University
• Keywords: spatial statistics ; cross covariogram ; covariogram

Abstract

Multivariate data are observed in many studies and often modeled by a multivariate spatial process. It is of practical interest to model the cross covariograms. Although parametric cross covariograms have been developed, inferences about the model parameters become more complex due to the increased number of parameters in the multivariate model. We provide a non-parametric alternative so that the cross covariograms introduce no additional parameters, thus facilitating simpler inferences and the handling of large multivariate data sets. We will compare the predictive performance of this non-parametric approach to some existing models.