Spatial prediction variance estimation based on covariance penalty

Résumé : In practice rarely (if ever) is the spatial covariance known in spatial prediction problems. Often, prediction is performed after estimated spatial covariance parameters are plugged into the prediction equation. The estimated spatial association parameters arealso plugged into the prediction variance of the spatial predictor. However, simply plugging in spatial covariance parameter estimates into the prediction variance of the spatial predictor does not take into account the uncertainty in the true values of the spatial covariance parameters. Therefore the plug-in prediction variance estimate will underestimate the true prediction variance of the estimated spatial predictor, especially for small datasets. We propose a new way to estimate the prediction variance of the estimated spatial predictor based on a covariance penalty using parametric bootstrapping. Our new estimator is compared to three other prediction variance estimators proposed in literature. The new prediction variance estimator generally performs better than the plug in method for small datasets with weak spatial association, although sometimes it is second best among the four prediction variance estimators compared. Furthermore, the new prediction variance estimator could potentially be used in the case of non-normal prediction.