Understanding the long-term trend of particulate phosphorus in a cyanobacteria-dominated lake using MODIS-Aqua observations

Information on the long-term trends in phosphorus (P) in lake waters is critical for clarifying transformation and biogeochemical cycling processes of P. We developed and validated an empirical model for deriving particulate phosphorus (PP, a dominant form of P) from MODIS-Aqua (Moderate Resolution Imaging Spectroradiometer) images. Subsequently, the long-term trend in PP in Lake Taihu from 2003 to 2017 and the driving factors were clarified. Based on the spectral index of the combination of remote sensing reflectance at 645 nm and 859 nm, a simple linear model was developed to derive PP for turbid cyanobacteria-dominated inland waters from MODIS-Aqua data (R-2 = 0.65; RMSE = 0.048 mg/L). Long-term MODIS observations show that PP demonstrated distinct spatial variations in Lake Taihu, with higher PP levels in cyanobacterial bloom-sensitive regions. There was a clear increasing trend in the PP of Lake Taihu, and the yearly average PP value increased from 0.089 mg/L in 2003 to 0.10 mg/L in 2017. A relatively strong positive correlation between 15-year spatially averaged PP data and algal bloom frequency revealed that cyanobacterial blooms mainly controlled the PP spatial variations. The daily average water temperature was significantly correlated with the daily PP derived from MODIS-Aqua data, indicating that the increasing water temperature was responsible for the PP increasing trend. An increase in water temperature can facilitate algal growth, thus resulting in a significant change in aquatic biogeochemical conditions, inducing more P release from the sediment and thereby increasing water PP levels. Our study has significant implications for understanding and controlling the P biogeochemical cycle in a cyanobacterial-dominated lake under the background of global warming. (C) 2020 Elsevier B.V. All rights reserved.