Combining soil maps with interpolations from point observations to predict quantitative soil properties

Abstract

A heuristic method is presented to predict quantitative soil properties by combining information from a soil map with information derived from additional point observations. This is done by taking a weighted average of soil map predictions and predictions obtained from kriging the observations. The weights are chosen such that the accuracies of the individual prediction methods are taken into account. A case study on the mapping of the mean highest water table in a Dutch polder area is used to test the method. Validation shows that combining the information from a 1:50,000 soil map with a kriged map produces a more accurate map than when either the soil map or the kriged map are used separately.