Modeling Skewness in Spatial Data Analyis without Data Transformation

Abstract

Skewness is present in a large variety of spatial data sets (rainfalls,winds, etc) but integrating such a skewness still remains a challenge.Classically, the original variables are transformed into a Gaussianvector. Besides the problem of choosing the adequate transform, thereare a few difficulties associated with this method. As an alternative,we propose a different way to introduce skewness. The skewness comesfrom the extension of the multivariate normal distribution to themultivariate skew-normal distribution. This strategy has many advantages.The spatial structure is still captured by the variogram and theclassical empirical variogram has a known moment generating function.To illustrate the applicability of such this new approach, we presenta variety of simulations.