RETRIEVING COLORED DISSOLVED ORGANIC MATTER USING S3-OLCI IN THE AMAZON RIVER CONTINUUM

The accurate retrieval of colored dissolved organic carbon (CDOM) information from remote sensing in optically complex waters is challenging. This study aimed to develop and evaluate an empirical algorithm (MBOWT) for retrieving CDOM in the Amazon River Continuum (ARC) using Sentinel-3 OLCI satellite data. Using in situ radiometry and CDOM samples collected over 14 years (2009–2023), the algorithm was designed to account for the optical water types of the study area. Validation with k-fold cross-validation (k = 20) demonstrated an overall agreement (median R² < 0.7). The algorithm was applied to S3-OLCI data processed with two atmospheric correction methods, Polymer and Acolite, with Acolite showing superior performance. Spatial analysis revealed distinct CDOM distribution patterns, highlighting higher concentrations in the northern Amazon River main channel compared to the southern branch and a marked decline before reaching the continental shelf break, demonstrating the algorithm’s potential for large-scale water quality monitoring.