Abstract: Remote sensing technology has emerged as a powerful tool for monitoring inland water quality. To investigate changes in concentrations of total phosphorus (TP), total nitrogen (TN), and chemical oxygen demand (COD) in the aquaculture area of Jianghan Plain, we developed inversion models utilizing Sentinel2-L1C remote sensing data. These models are based on three optical active substances: water transparency, suspended matter concentration, and Chl-a concentration. By correlating the calculated water transparency, suspended solids concentration, and Chl-a concentration with TP, TN, and COD concentrations at regional water fitting points, we established indirect inversion models to estimate the concentrations of these substances and validated the models using fitting points data. Our results demonstrate a strong correlation between water transparency and TP concentration. Higher water transparency indicates better water quality and lower TP concentration. The concentration of suspended matter in water is highly correlated with the TN concentration. Increased suspended matter concentration indicates inferior water quality and higher TN concentration. Additionally, higher water Chl-a concentration corresponds to higher COD. The deterministic coefficients of the indirect fitting models for TP concentration, TN concentration, and COD were all greater than 0.6. Furthermore, the TP concentration, TN concentration, and COD concentration in water bodies of monitoring points in Jianghan Plain were simulated. The results align with the expected patterns in feeding period and reflect the spatio-temporal evolution of water quality parameters in the aquaculture areas of Jianghan Plain. This research contributes practical insights into understanding the dynamics of large-scale water quality parameters in aquaculture areas.

Key words: water quality parameters, Sentinel2-L1C image, Jianghan Plain, optically active compounds, retrieval model