Long-term remote observations of particulate organic phosphorus concentration in eutrophic Lake Taihu based on a novel algorithm.

Monitoring the long-term spatiotemporal variations in particulate organic phosphorus concentration (C) is imperative for clarifying the phosphorus cycle and its biogeochemical behavior in waters. However, little attention has been devoted to this owing to a lack of suitable bio-optical algorithms that allow the application of remote sensing data. In this study, based on Moderate Resolution Imaging Spectroradiometer (MODIS) data, a novel absorption-based algorithm of C was developed for eutrophic Lake Taihu, China. The algorithm yielded a promising performance with a mean absolute percentage error of 27.75% and root mean square error of 21.09 μg/L. The long-term MODIS-derived C demonstrated an overall increasing pattern over the past 19 years (2003-2021) and a significant temporal heterogeneity in Lake Taihu, with higher value in summer (81.97 ± 3.81 μg/L) and autumn (82.07 ± 3.8 μg/L), and lower C in spring (79.52 ± 3.81 μg/L) and winter (78.74 ± 3.8 μg/L). Spatially, relatively higher C was observed in the Zhushan Bay (85.87 ± 7.5 μg/L), whereas the lower value was observed in the Xukou Bay (78.95 ± 3.48 μg/L). In addition, significant correlations (r > 0.6, P < 0.05) were observed between C and air temperature, chlorophyll-a concentration and cyanobacterial blooms areas, demonstrating that C was greatly influenced by air temperature and algal metabolism. This study provides the first record of the spatial-temporal characteristics of C in Lake Taihu over the past 19 years, and the C results and regulatory factors analyses could provide valuable insights for aquatic ecosystem conservation.