Monitoring phycocyanin concentrations in high-latitude inland lakes using Sentinel-3 OLCI data: The case of Lake Hulun, China

Abstract

With the intensification of global warming, eutrophication in lakes at high latitudes of China has become increasingly severe, with the harm of blue-green algae blooms also on the rise. Therefore, it is urgent to conduct research on water quality of lakes in high latitudes. In this study, taking Lake Hulun as an example, a phycocyanin (PC) inversion model applicable to Sentinel-3 OLCI data was constructed and applied to the Sentinel-3 dataset from 2016 to 2022 to analyze the spatiotemporal variation characteristics of PC concentration. The driving mechanism of climate factors on PC concentration was explored, and the correlation between PC concentration and Cyanobacterial blooms (CBs) outbreak was analyzed. Results showed that the PC concentration inversion model based on XGBoost (XGB) has the highest accuracy (R2 = 0.91, RMSE = 76.76 μg/L, and rRMSE = 0.54). Monthly average PC concentration is higher in July (44.52±64.85 μg/L) and lower in October (5.04±1.81 μg/L). From 2016 to 2022, the annual average concentration of PC in Hulun Lake in 2022 (38.82±63.34 μg/L) is higher than that in other years, while the annual average PC concentration in 2020 (4.60±1.76 μg/L) is lower. Temperature is the main impacting factor on PC concentration. The variation of PC concentration in Lake Hulun has high spatiotemporal consistency with the proportion of CBs area. In summary, using Sentinel-3 OLCI imagery for long-term remote sensing monitoring of spatiotemporal pattern changes of PC in Lake Hulun, and analyzing its changing characteristics and patterns, is of great significance for early warning of CBs.