Remote sensing and in situ observations of harmful algal blooms in dongzhang reservoir: implications for water quality management
摘要
Harmful Algal Blooms (HABs) cause significant ecological damage and public health issues in freshwater reservoirs globally. While in situ sampling is essential for monitoring HABs, it is often costly and time-consuming. Remote sensing offers a rapid and cost-effective alternative for detecting HABs. This study integrates Sentinel-2 multispectral imagery with in-situ observations to monitor HABs in Dongzhang Reservoir, Fujian, China. We analyzed the Normalized Difference Vegetation Index and Normalized Difference Chlorophyll Index (NDCI) to monitor chlorophyll-a concentrations in this dynamic water body. In situ water samples were collected at three locations to determine the phytoplankton density, and these counts were correlated with spectral indices. Our findings indicate that the Dongzhang Reservoir experiences high algal biomass during spring and summer, as evidenced by seasonal variations in spectral index values. The NDCI, which leverages the red edge wavelength, provided the most accurate identification of chlorophyll-a. Linear regression analysis revealed strong correlations between phytoplankton density and remote sensing values in spring, summer, and autumn of 2022, with R2 values of 0.81, 0.74, and 0.62, respectively. Two key factors driving HAB proliferation in 2022 were total phosphorus levels and water surface temperatures, which were positively correlated with phytoplankton density (R2 = 0.83 for total phosphorus and R2 = 0.81 for water temperature). In 2023, total phosphorus remained the main factor, with an R2 of 0.75. HAB events were more severe in 2022, affecting 34.41% of the reservoir area in May, compared with 22.15% in 2023. This study demonstrated that Sentinel-2 imagery captures greater spatial and temporal heterogeneity of algal blooms than traditional in situ samples do, highlighting the potential of remote sensing as a crucial tool for monitoring algal bloom dynamics in reservoirs and other water bodies. Complementing in situ samples with Sentinel-2 images, owing to their high temporal resolution, can increase monitoring efforts. The implementation of appropriate management strategies can mitigate HABs in Dongzhang Reservoir.