An empirical ocean colour algorithm for estimating the contribution of coloured dissolved organic matter in North-Central Western Adriatic Sea

The performance of empirical band ratio models were evaluated for the estimation of Coloured Dissolved Organic Matter (CDOM) using MODIS ocean colour sensor images and data collected on the North-Central Western Adriatic Sea (Mediterranean Sea). Relationships between in situ measurements (2013-2016) of CDOM absorption coefficients at 355 nm (aCDOM355) with several MODIS satellite band ratios were evaluated on a test data set. The prediction capability of the different linear models was assessed on a validation data set. Based on some statistical diagnostic parameters (R2, APD and RMSE), the best MODIS band ratio performance in retrieving CDOM was obtained by a simple linear model of the transformed dependent variable using the remote sensing reflectance band ratio Rrs(667)/Rrs(488) as the only independent variable. The best-retrieved CDOM algorithm provides very good results for the complex coastal area along the North-Central Western Adriatic Sea where the Po River outflow is the main driving force in CDOM and nutrient circulation, which in winter mostly remains confined to a coastal boundary layer, whereas in summer it spreads to the open sea as well.