Assessing how heavy metal pollution and human activity are related by using logistic regression and kriging methods

Quantifying how soil pollution and a long-term land-use perspective (i.e. human activity) are related is a highly effective means of managing soil resources in central Taiwan. By defining hazard zone as the heavy metal contents that exceed the corresponding control standard, this study estimates not only the spatial patterns of hazardous probability based on only the observed heavy metal data using indicator kriging (IK), but also those that consider auxiliary variables by logistic regression (LR) and regression kriging (RK). Estimation results indicate that the hazard pattern estimated by the IK and RK is more fragmented than those estimated by LR. Moreover, the LR and RK, can determine how a pollution source and a pathway are related. Based on the results, the hazard area is strongly correlated with the locations of industrial plants and irrigation systems in the study area. These methods provide an effective means of exploring hazard risks efficiently for future monitoring of soil contamination. The LR and RK methods not only identify natural and human factors of soil pollutions, but also enhance delineations of identifying hazardous area of soil pollution. Particularly, the RK considers the spatial residuals to improve the goodness of fit in the LR.

Research highlightsThis study detects the relationships between pollution and human activity for a specific case study.
► The methods identify pollution driving factors and enhance delineations of hazardous area.
► Driving factors such as distances to channel and factory are strongly correlated to soil pollution.
► Soil pollution is closely related to pollution sources such as industrial factories in study area.
► The models are highly promising for use in evaluating heavy metal susceptibility to individuals.