Long-term Observations of Chlorophyll-a Concentration in Lake Honghu Using Multi-Source Remote Sensing Data

Zhe Chen,Yu Xia,Ying Jiang,Jing Zhao,Yibang Wu,Jingwei Li

DOI: https://doi.org/10.1117/12.2597710

2021-01-01

Abstract:As an important area of wetland protection in the Yangtze River Economic Belt, Honghu Lake, is a habitat or wintering place for many wild animals. In the past two decades, the water environment of Honghu Lake has been suffering from serious deterioration due to the rapid development of agriculture and industry in the urbanization of surrounding areas. Under the combined effect of multiple non-point source pollution and point source pollution, cyanobacteria blooms often occur here. As a proxy of eutrophication, chlorophyll-a (Chl-a) has been considered to be an important indicator of water quality parameters. For better understanding of the change of the water quality of Honghu Lake, an improved empirical model for estimating chlorophyll-a concentrations (Chl-a) from different multi-spectral satellites images was established and validated. (1) We combined the results of two-band algorithms(2BDA), three-band algorithms(3BDA), normalized difference chlorophyll index (NDCI) and fluorescence line height (FLH) into support vector machine model (SVM) for better multi-nonlinear relationship establishment between Chl-a concentration and surface water reflectance, which acquired higher model accuracy.(2) Based on the long-term time series data derived from Landsat-7, Landsat-8 and Sentinel-2, the variation of Chl-a concentration of Honghu Lake over long term was obtained. (3) Our results demonstrate that the average chlorophyll concentration has been at a very high level and showing an increasing tendency in recent years, which may indicate the eutrophication in the Honghu Lake is still getting worse.