Transparency inversion algorithm for surface drinking water sources based on high-resolution satellite (GF-2) data

Global warming, human activities, and extreme weather events have exacerbated issues such as freshwater scarcity and water pollution, posing significant threats and challenges to the safety of surface drinking water sources.Accurate and rapid monitoring of water quality of surface drinking water sources are necessary to ensure drinking water safety. Water transparency is an important indicator for evaluating water quality and plays an important role in the management of surface drinking water sources. The traditional method of obtaining transparency is to use a Secchi disk (SD) for measurement, but the data obtained are spatially limited and discontinuous, and are not suitable for large-scale water body monitoring. Satellite remote sensing technology has become an effective method for quickly obtaining water parameters due to its advantages in acquiring large-scale, realtime data, as well as in saving on labor costs and material resources, and overall, to improve the efficiency of the work. Especially for small and medium-sized surface water sources, high-resolution satellite data can provide more refined and accurate monitoring information, so as to more effectively manage and protect these water sources. In view of the existing water transparency inversion algorithms are mostly based on ocean satellite data with low spatial resolution, this study utilizes China’s first civilian sub-meterlevel GF-2 (Gaofen2) satellite images to perform the transparency observation for the reservoirs in Hainan Island, China. The high spatial and temporal resolutions of GF-2 facilitate the potential for the monitoring of water quality of surface drinking water sources, especially for island reservoirs, which are often small to medium-sized. However, how to overcome the limitations of the limited multi-spectral bands setting of the GF-2 satellite data and to enhance its capability in extracting the information of the surface water is a challenge. Based on in-situ measurement of Secci transparency and reflectance spectrum of water surface, at least three achievements were gained in this study: 1) a water transparency inversion algorithm was constructed based on the empirical regression method, and the accuracy and suitability of the model were examined to ensure its applicability to different water bodies. 2) The algorithm was applied to the GF-2 image, the transparency of representative reservoirs in Hainan Province was inverted, and the feasibility of the inversion algorithm for GF2 was confirmed. Transparency distribution maps were drawn and the spatial and temporal distribution characteristics of the transparency of representative reservoirs in Hainan Island were analyzed. 3) based on the in-situ measurements of chlorophyll a and suspended sediments, the causes inducing the seasonal and annual change of transparency were discussed. The study provides technical support for the management and protection of surface drinking water, especially for small and medium-sized reservoirs.