Integration of categorical information of land use maps in spatial prediction of soil available Cu in Hanchuan county, China

Predicting soil available Cu content distribution accurately from limited soil samples has received a great deal of attention recently in an effort to understand the status of nutrients or pollutants in soils. Map quality can affect the effectiveness of site-specific management. In this study, the point data of available Cu contents at 402 sampling sites and the areal data based on the land use map were integrated together for performing area and point kriging (AAPK). A comparison in the performances of AAPK and ordinary kriging (OK) was conducted based on 135 validation samples to assess the feasibility of AAPK in improving the interpolation quality of soil available Cu contents. Results indicate that the measured data at validation locations are more correlated with predicted data by AAPK than with those by OK, and the root mean square error with AAPK is lower than that with OK. Furthermore, it can be found from the prediction variance maps that AAPK always generates smaller error variances than OK. It can be concluded that land use type information has important impact on the spatial distribution of soil available Cu, and AAPK is an effective and practical prediction method for increasing the interpolation accuracy of soil available Cu.