Water quality monitoring of large reservoirs in China based on water color change from 1999 to 2021

Abstract

Water color is an essential indicator of water quality, and research methods of water color based on remote sensing have become a trend in large-scale water quality monitoring. Large reservoirs are important for regional development. However, large-scale research into the quality of the water is lacking. Accordingly, to study the spatiotemporal water quality conditions of large reservoirs in China, this study conducted an inversion of water color over the past 23 years. Using Landsat series images, the Dynamic Surface Water Extent (DSWE) algorithm, and the Forel-Ule index (FUI), the water surface of 746 large reservoirs in China was sampled and the water color was inverted using the Google Earth Engine (GEE) platform. The FUI local change index was established to analyze changes over time. The results showed that the annual average FUI value of large reservoirs in China was less than 11, showing non-eutrophic and green water bodies. From the water color analysis of 91 large Type I reservoirs, it was found that the water quality of the reservoirs in the Northeast region was poor (FUI greater than 12), with many aquatic plants and algae or highly turbid and greenish-brown waters, mainly for flood control and irrigation. In southern China, there were primarily mesotrophic and oligotrophic reservoirs (FUI not greater than 8) with green or greenish-blue water color, which were mainly used for water supply and power generation. The construction of the southern reservoirs reduced the FUI values by an average of five levels, apparently changing the water color. The FUI local change index revealed that the FUI values of large reservoirs in China showed an overall decreasing trend from 1999 to 2007, stabilized after 2007, and then fluctuated after 2013. This study also found that the water color index of large reservoirs in China is positively correlated with the trophic status and turbidity of water bodies to some extent. Therefore, this study can compensate for the lack of water color remote sensing in large-scale reservoir water quality studies and support large-scale reservoir monitoring and management.