A novel dynamic pollution risk assessment method for potential pollution areas in industrial parks: Integrating...

Soil and groundwater pollution in industrial parks arises from multiple pollution sources and overlapping spatiotemporal risk processes, posing a persistent threat to human health. Integrating remote sensing imagery with multi-source data is a promising approach for quantifying and mapping pollution risk distributions. However, the potential of high-resolution historical remote sensing imagery to capture long-term dynamic indicators at critical facility scale remains underutilized. This study develops a dynamic pollution risk assessment framework within a source-pathway-receptor paradigm, incorporating both static and dynamic indicators. A key innovation is the integration of long-term sub-meter remote sensing imagery, which enables the synchronous delineation of potential pollution areas and extraction of remote-sensing-based dynamic (RSD) indicators, such as surface rust, construction activities, and vegetation cover dynamics. The proposed method was applied to evaluate both current (2024) and historical cumulative (2005-2024) risks in a typical petrochemical industrial park in the Beijing-Tianjin-Hebei region. Correlation analysis and Light Gradient Boosting Machine (Light GBM) regression were employed to investigate the temporal variations in key drivers. Results indicated that cumulative effects increased overall pollution risk by 23.4%. Although the composition of pollution risks exhibited spatiotemporal heterogeneity, current risk was predominantly governed by chemical inventory and equipment age, while the contribution of dynamic indicators increased markedly with historical risk accumulation. This dynamic driving mechanism highlights the necessity of simultaneously controlling current risks and mitigating cumulative risks. The proposed methodology enables comprehensive and temporally explicit evaluation of pollution risks, supporting the transition of industrial parks from end-of-pipe remediation toward proactive, source-oriented risk prevention.