MEMBER

Optimization of Sampling Strategies for Pollution Source Identification in River Water Quality Monitoring

Early detection of water pollution and identification of pollution sources are essential for effective management of aquatic environments. However, dense monitoring of entire river watersheds is difficult because water sampling and chemical analysis require substantial labor and cost. Therefore, sampling locations must be selected strategically to obtain useful information from limited observations. This research aims to develop a sampling site selection method for pollution source identification, targeting watershed pollutants such as PFAS and heavy metals. To evaluate candidate sampling sites, concentration responses under different source locations and release conditions must be predicted in advance. However, many pollutants undergo environmentally dependent changes in chemical form, solubility, and adsorption/desorption behavior, making watershed-scale simulation challenging. This research will establish a water quality simulation framework that couples chemical reaction processes with physical transport processes. The framework will be used to generate concentration time-series data under diverse source locations and release conditions. The resulting dataset will then be used to identify combinations of sampling sites that are effective for pollution source identification and anomaly detection. This approach will support the rational design of monitoring plans under limited observational resources and contribute to more advanced river water quality management.