A hybrid remote sensing approach for estimating chemical oxygen demand concentration in optically complex waters: A case study in inland lake waters in eastern China
Xiaolan Cai,Yunmei Li,Shaohua Lei,Shuai Zeng,Zhilong Zhao,Heng Lyu,Xianzhang Dong,Junda Li,Huaijing Wang,Jie Xu,Yuxin Zhu,Luyao Wu,Xin Cheng
DOI: https://doi.org/10.1016/j.scitotenv.2022.158869
2022-10-05
Abstract:Chemical oxygen demand concentration (C COD ) is widely used to indicate the degree of organic pollution of lakes, reservoirs and rivers. Mastering the spatiotemporal distribution of C COD 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 COD in inland optically complex waters embedding the interaction between C COD 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 2 = 0.81), Lake Qiandaohu (MAPE = 21.63 %, RMSE = 0.50 mg/L and R 2 = 0.69), and Yangtze River (MAPE = 29.34 %, RMSE = 0.83 mg/L, and R 2 = 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 COD 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 COD in optically complex waters, and that satellite-derived products can provide reliable information for lake water quality management.
environmental sciences