Remotely sensed Coloured Dissolved Organic Matter distribution in the Amazon River Continuum

Coloured Dissolved Organic Matter (CDOM) plays a key role in aquatic carbon cycling, light availability and ecosystem processes. Accurate retrieval of CDOM absorption in optically complex waters, such as those of the Amazon River, remains challenging due to overlapping absorption features with non-algal particles. This study evaluates existing empirical and semi-analytical algorithms for retrieving CDOM absorption at 412 nm using in situ data from the Amazon River Continuum and develops two new empirical models. Incorporating wavelengths longer than those typically used, particularly 674 nm, significantly improves retrieval performance. Two novel models were developed: the Band-Ratio Optical Water Type (BROWT), tailored to optical water types within the study area, and the Band-Ratio General Sensor (BRGS), designed for cross-sensor applicability. BROWT achieved superior in situ performance (R2 = 0.87, MRAD = 40%) and outperformed all evaluated literature algorithms. In satellite match-up analyses, both models showed strong agreement with in situ observations. Acolite and Polymer atmospheric correction processors yielded comparable accuracy, each involving trade-offs between retrieval precision and spatial coverage. Applied to Sentinel-3 OLCI imagery, BROWT revealed pronounced seasonal and spatial variability in aCDOM412 across the Amazon River Continuum. The South Channel consistently exhibited higher aCDOM412 than the North Channel (mean: 2.13 vs. 1.98 m−1 in 2022; 2.07 vs. 1.97 m−1 in 2023), while the inner plume showed mean values of 2.28 m−1 (2022) and 2.16 m−1 (2023). Effect size analysis (Cohen’s d = 0.02–0.30) identified the South Channel as the dominant CDOM pathway to the offshore plume. Elevated aCDOM412 during the High and Falling water seasons, with Falling water exceeding High water despite comparable discharge, is consistent with sustained floodplain organic matter inputs. These results demonstrate the utility of satellite remote sensing for monitoring CDOM dynamics and carbon fluxes in large, heterogeneous river systems.