Assessment of heavy metal concentrations and pollution indices using geographically weighted regression: A case study in a Hyrcanian forest, Northern Iran

: Heavy metal contamination of surface soils has become a significant environmental concern in many watersheds due to rapid urbanization, industrial development, and changes in land use and land cover (LULC). This study evaluates the concentrations, spatial distribution, and ecological risks of selected heavy metals in the surface soils of the Vaz watershed in Mazandaran Province, northern Iran. This area is of high ecological value and raises public health concerns because of extensive rice cultivation and its direct connection to the Caspian Sea. Soil samples were analysed to determine heavy metal concentrations across LULC classes. Contamination status and ecological implications were assessed using the Geo-accumulation Index (Igeo), Pollution Load Index (PLI), and Potential Ecological Risk Index (PERI). Additionally, Geographically Weighted Regression (GWR) was used to explore spatially varying relationships between heavy metal concentrations and explanatory variables, particularly LULC patterns. The results reveal significant spatial variability in heavy metal concentrations, with higher levels generally found in agricultural and residential areas, indicating the influence of human activities. The ecological risk indices suggest that certain locations within the watershed pose moderate to considerable ecological risks. The GWR analysis highlights significant spatial non-stationarity, demonstrating that the effects of LULC on heavy metal accumulation vary across the watershed. Given the importance of the Vaz River in transporting pollutants to the Caspian Sea and the reliance of local communities on agricultural products, these findings provide a comprehensive framework and essential insights for environmental management, land-use planning, and risk mitigation to support sustainable watershed management.