Improved heavy metal mapping and pollution source apportionment in Shanghai City soils using auxiliary information

Soil heavy metal pollution can be a serious threat to human health and the environment. The accurate mapping of the spatial distribution of soil heavy metal pollutant concentrations enables the detection of high pollution areas and facilitates pollution source apportionment and control. To make full use of auxiliary soil properties information and show that they can improve mapping, a synthesis of the Bayesian Maximum Entropy (BME) theory and the Geographically Weighted Regression (GWR) model is proposed and implemented in the study of the Shanghai City soils (China). The results showed that, compared to traditional techniques, the proposed BME-GWR synthesis has certain important advantages: (a) it integrates heavy metal measurements and auxiliary information on a sound theoretical basis, and (b) it performs better in terms of both prediction accuracy and implementation flexibility (including the assimilation of multiple data sources). Based on the heavy metal concentration maps generated by BME-GWR, we found that the As, Cr and Pb concentration levels are high in the eastern part of Shanghai, whereas high Cd concentration levels were observed in the northwestern part of the city. Organic carbon and pH were significantly correlated with most of the heavy metals in Shanghai soils. We concluded that Cd pollution is mainly the result of agricultural activities, and that the Cr pollution is attributed to natural sources, whereas Pb and As have compound pollution sources. Future studies should investigate the implementation of BME-GWR in the case of space-time heavy metal mapping and its ability to integrate human activity information and soil category variables.