The authors propose a method for fitting regression models to data that exhibit spatial correlation and heteroskedasticity. A combination of parametric and nonparametric regression techniques is used to iteratively estimate the various components of the model. The approach is demonstrated on a large dataset of predicted nitrogen runoff statistics from agricultural land in the Midwest and Northern Plains.
This working paper was published as J. D. Opsomer, D. Ruppert, M. P. Wand, U. Holst and O. Hossjer, "Kriging with Nonparametric Variance Function Estimation," Biometrics 55 (1999): 704–710.
Opsomer, Jean D.; Ruppert, D.; Wand, M. P.; Holst, U.; and Hössjer, O., "Kriging With Nonparametric Variance Function Estimation" (1998). CARD Working Papers. 226.