Exploring spatial distributions of increments in soil heavy metals and their relationships with environmental factors using GWR

Due to human activities and industrial production, heavy metals accumulate continuously in soils, resulting in environmental ecological risks. Thus, it is critical to reveal the spatial patterns of the increments in soil heavy metals and their influencing factors to prevent the continuous deterioration of soil due to heavy metal pollution. In this study, based on soil samples collected in 2016 and 2019 at the same sites in the southern part of Daye city, the spatial distributions of increments in soil heavy metals were obtained using spatial interpolation and overlap methods. Then, the geographically weighted regression (GWR) model was used to analyze the influence of various environmental factors in three categories (location characteristics, topographical factors, and soil properties) on the increments in soil heavy metals. The results showed the following: (1) The soils in the study region were severely polluted with Cd, Cu, Pb, and Zn. Throughout almost the whole study region, the concentrations of these four heavy metals in soil exceeded local background values. (2) The concentrations of Cd, Cu, Pb, and Zn increased from 2016 to 2019 in 77.38%, 59.71%, 68.42%, and 49.21% of the study region, respectively. According to the spatial distribution of comprehensive change index values, soil heavy metal pollution continued to deteriorate in 74.4% of the study region from 2016 to 2019. (3) The GWR model revealed spatially varying relationships between the increases in soil heavy meals and environmental factors, and the results indicated that location characteristics and topographical factors had the largest and smallest influences, respectively, on the spatiotemporal increments in soil heavy metals. The influences of soil properties on the increments in soil heavy metals were similar to the influences on their concentrations. The GWR model had a higher R 2 and lower AICc than the ordinary least square regression model, indicating that GWR had a stronger ability to explain the relationships between the increments in soil heavy metals and environmental factors.