Methods to correct and compute confidence and prediction intervals of models neglecting sub-parameterization heterogeneity - From the ideal toward practice
We derive methods to correct regression-based confidence and prediction intervals. The corrections compensate for sub-parameterization heterogeneity. The correction methods require from ideal to little geostatistical characterization. Weighting can be estimated as part of the regression and the interval calculation. Testing demonstrates the importance and accuracy of the interval corrections.