Simultaneous retrieval of chlorophyll-a (Chla) and total suspended solids (TSS) from Landsat-8, Sentinel-2, and Sentinel-3

Abstract:

Biogeochemical properties, such as chlorophyll-a (Chla) and total suspended solids (TSS), are critical proxies for understanding the status, health, trophic state, and carbon budget in aquatic ecosystems. Remote sensing plays an important role in bridging the spatial and temporal gaps in knowledge of the distribution of Chla and TSS across vast geographic scales. Here, we formulate a machine-learning model to simultaneously retrieve these two biogeochemical indicators from in situ radiometric spectra. We anticipate our simultaneous retrievals, trained and validated with a global dataset, to yield major improvements over the state-of-the-art algorithms. The model is applicable to atmospherically corrected Landsat-8, Sentinel-2, and Sentinel-3 optical imagery, enabling the production of multimission spatio-temporal Chla and TSS maps.