An approach for mapping phytoplankton communities in freshwater lakes based on phytoplankton absorption features.

Phytoplankton communities play a crucial role in the lake ecosystem due to their varying characteristics, functions, and impacts of different phytoplankton groups. Understanding the composition of phytoplankton groups in freshwater lakes is essential for comprehending geochemical processes and managing water quality. In this study, an improved Diagnostic Pigment Analysis method for freshwater lakes was developed and the proportion of five major phytoplankton groups-Dinophyta, Cryptophyta, Chlorophyta, Cyanophyta, and Bacillariophyta-was derived through the absorption-decomposition method. The validation results demonstrated that the developed algorithm had satisfactory estimation accuracy for all five groups. Among all the phytoplankton groups, Cyanophyta achieved the best performance, with Median Absolute Percentage Error (MAPE) of 14.22 %, and Bias of 8.37 %. In contrast, Cryptophyta exhibited the poorest accuracy, with MAPE as high as 40.24 %. The MAPE values ranged from 10.91 % to 33.65 %, and the Bias values ranged from 1.06 % to 9.38 %. Meanwhile, the developed algorithm was successfully applied to the Ocean and Land Color Instrument (OLCI) images for mapping the spatial distribution of phytoplankton communities in Lake Taihu, demonstrating its ability to be applied to satellite imagery. This proposed algorithm provided a new approach to quantitatively determine the composition of phytoplankton communities in freshwater lakes, which can obtain valuable insights from observing the composition and succession patterns of these communities from satellite platforms.