Xianzhang Dong

Recently, numerous bio-optical algorithms have been proposed to estimate chlorophyll-a (Chla) concentration in global surface waters. The surge of algorithms is mainly due to the heterogeneity of waters, especially in inland systems. To reduce redundant modeling work for inland waters and to enhance the usability of algorithms, it is urgent to evaluate the published algorithms for similar optical water types. However, approaches of the algorithm assessment in previous studies vary in error metrics usage or relative comparison, leading to the current lack of a standard or consensus in the community. Thus, we proposed a round-robin scoring method based on optically fuzzy clustering for algorithm assessment. An improved scoring approach (namely the sort-based method) was employed to aggregate the metrics of candidate algorithms. We generate a new set of optical water types using 1702 in situ samples, showing an improvement on our previous work [Bi et al. (2019)]. Then, an integrated Chla result was produced by blending the estimations of optimal algorithms weighted by the membership values, which showed better performance than that of any single algorithm. The blending framework presents more flexible and expandable than previous studies. We aim to illustrate the importance of assessment which considers key properties of algorithms (i.e., Accuracy , Precision , and Effectiveness ) and to advocate more in-depth research on certain water types under the fuzzy clustering-based estimation framework (such as clean inland waters with low Chla). This article is accompanied by the open-source R package FCMm for implementation of the proposed method.

Due to eutrophication and water quality deterioration in clear reservoirs, it is necessary to monitor and manage the main water parameters: concentration of total phosphorus (C_TP), chemical oxygen demand (C_COD), chlorophyll-a (C_Chla), total suspended matter (C_TSM), and Secchi disk depth (SDD). Five random forest (RF) models are developed to estimate these parameters in Xin’anjiang Reservoir, which is a clear drinking water resource in Zhejiang, China. Then, the spatio-temporal distributions of the parameters over 7 years (2013–2019) are mapped using GaoFen-1 (GF-1) images and the relationships with driving factors are analyzed. Our study demonstrates that the parameters’ distributions exhibited a significant spatio-temporal difference in Xin’anjiang Reservoir. Spatially, relatively high C_TP, C_COD, C_Chla, and C_TSM but low SDD appear in riverine areas, showing strong evidence of impact from the incoming rivers. Temporally, C_Chla and C_TSM reached high values in summer and winter, whereas SDD and C_TP were higher in the summer and autumn, respectively. In contrast, no significant seasonal variations of C_COD could be observed. This may be why C_COD is not sensitive to hydrological or meteorological factors. However, precipitation had a significant impact on C_Chla, C_TP, SDD, and C_TSM in riverine areas, though these parameters were less sensitive to meteorological factors. Moreover, the geomorphology of the reservoir and anthropogenic interference (e.g., tourism activities) also have a significant impact on the water quality parameters. This study demonstrates that coupling long-term GF-1 images and RF models could provide strong evidence and new insights to understand long-term dynamics in water quality and therefore support the development of corresponding management strategies for freshwater reservoirs.

The accurate remote estimation of the Secchi disk depth (Z_SD) in turbid waters is essential in the monitoring the ecological environment of lakes. Using the field measured Z_SD and the remote sensing reflectance (Rrs(λ)) data, a new semi-analytical algorithm (denoted as Z_SDZ) for retrieving Z_SD was developed from Rrs(λ), and it was applied to Geostationary Ocean Color Imager (GOCI) images in extremely turbid waters. Our results are as follows: (1) the Z_SDZ performs well in estimating Z_SD in turbid water bodies (0.15 m < Z_SD < 2.5 m). By validating with the field measured data that were collected in four turbid inland lakes, the determination coefficient (R²) is determined to be 0.89, with a mean absolute square percentage error (MAPE) of 22.39%, and root mean square error (RMSE) of 0.24 m. (2) The Z_SDZ improved the retrieval accuracy of Z_SD in turbid waters and outperformed the existing semi-analytical schemes. (3) The developed algorithm and GOCI data are in order to map the hourly variation of Z_SD in turbid inland waters, the GOCI-derived results reveal a significant spatiotemporal variation in our study region, which are significantly driven by wind forcing. This study can provide a new approach for estimating water transparency in turbid waters, offering important support for the management of inland waters.