Sentinel-3 OLCI observations of water clarity in large lakes in eastern China: Implications for SDG 6.3.2 evaluation
The Sustainable Development Goal (SDG) 6.3.2 of the United Nations (UN) focuses on ambient water quality, while water clarity simplistically and visually reflect water quality and can potentially support SDG 6.3.2 reporting. In this study, based on extensive field data and Sentinel-3 Ocean and Land Color Instrument (OLCI) imagery, a random forest regression (RFR) Secchi depth (Zsd) model suitable for turbid and eutrophic waters was established. With this model, the Zsd of 86 large (> 30 km2) lakes in Eastern China was obtained from May 2016 to April 2018. Additionally, the potential for applying OLCI-derived Zsd data in the SDG 6.3.2 evaluation was assessed. Of six common atmospheric correction (AC) processors (i.e., BAC, C2RCC, POLYMER, BP, MUMM, and 6SV), 6SV often exhibited the best performance except for at 754 nm (root mean square error (RMSE) ≤ 0.0094 sr-1, mean absolute percentage error (MAPE) ≤ 36.27%, and mean normalized bias (MNB) ≤ 15.89%). The RFR model had higher accuracy (R2= 0.70, RMSE = 0.13 m, MAPE = 33.43%, and MNB = 14.55%) and was more suitable for eutrophic and turbid inland lakes across Eastern China than the other available Zsd algorithms. The average OLCI-derived Zsd of the lakes in Eastern China was 0.44 ± 0.13 m, suggesting that these lakes are extremely turbid. The average Zsd of lakes in the Eastern Plain Lake (EPL) zone (0.45 ± 0.12 m) was higher than that of the lakes in the Northeastern Plain and Mountain Lake (NPML) zone (0.40 ± 0.17 m). The majority of lakes showed higher Zsd values in summer (rainy season) than in fall and winter (dry season). A simple SDG 6.3.2 evaluation scheme was developed based on the Zsd product, and only 54.65% of the lakes (N = 47) reached the “good” level during the monitoring period as a result of the eutrophication of the lakes in Eastern China. This study provides water clarity information for large lakes in Eastern China and facilitates the understanding of ambient water quality under the 2030 UN SDG framework as well as data and technical support for future SDG 6.3.2 evaluations.
作者:
Ming Shen, Hongtao Duan et al. Remote Sensing of Environment. 2020. DOI: https://doi.org/10.1016/j.rse.2020.111950
