A. GARCIA PEREZ
Let Z(s) be a multivariate spatial process. If we have a sample of Z(s) at n locations, we measure the statistical association between the random components of Z with the correlation coefficient and the spatial dependence with their variograms. If two of the omponents are correlated, the spatial information provided by one of them can be improved by the other one. To capture this association both within components of Z(s) and across s, we use the cross-variogram.
Only two robust cross-variogram estimators have been proposed in the literature by Lark (2003) and moreover, their sample distributions were not obtained.
In this paper we proposed new robust cross-variogram estimators, following the location estimation way proposed in García-Pérez (2020, 2021), instead of the scale estimation considered by Lark, and we obtain a saddlepoint approximation for their sample distributions, assuming a multivariate scale contaminated normal model.
Keywords: Robustness, Spatial data, Saddlepoint approximations
Scheduled
GT04 Multivariate Analysis and Classification I
June 7, 2022 12:00 PM
Cloister room