Chemical oxygen demand concentration (C) is widely used to indicate the degree of organic pollution of lakes, reservoirs and rivers. Mastering the spatiotemporal distribution of C is imperative for understanding the variation mechanism and controlling of organic pollution in water. In this study, a hybrid approach suitable for Sentinel 3A/Ocean and Land Colour Instrument (OLCI) data was developed to estimate C in inland optically complex waters embedding the interaction between C and the absorption coefficients of optically active constituents (OACs). Based on in-situ sampling in different waters, the independent validations of the proposed model performed satisfactorily in Lake Taihu (MAPE = 23.52 %, RMSE = 0.95 mg/L, and R = 0.81), Lake Qiandaohu (MAPE = 21.63 %, RMSE = 0.50 mg/L and R = 0.69), and Yangtze River (MAPE = 29.34 %, RMSE = 0.83 mg/L, and R = 0.64). In addition, the approach not only showed significant superiority compared with previous algorithms, but also was suitable for other common satellite sensors equipped same or similar bands. The hybrid approach was applied to OLCI images to retrieve C of Lake Taihu from 2016 to 2020 and reveals substantial interannual and seasonal variations. The above results indicate that the proposed approach is effective and stable for studying spatiotemporal dynamic of C in optically complex waters, and that satellite-derived products can provide reliable information for lake water quality management.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1016/j.scitotenv.2022.158869 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!