RESUMEN
For the last two decades different scientific disciplines have focused on lacustrine dissolved organic matter (DOM) given its importance in the biogeochemistry of carbon and in ecosystem functioning. New satellites supply the appropriate resolutions to evaluate chromophoric dissolved organic matter (CDOM) in inland waters, opening the possibility to estimate DOM at appropriate spatiotemporal scales. This requires, however, a robust relationship between CDOM and dissolved organic carbon (DOC). In this work, we evaluated the use of CDOM as a proxy of DOC in 7 Andean Patagonian lakes. Considering the entire data set, CDOM absorption coefficients (a355 and a440) were linearly related with DOC. Shallow lakes, however, drove this relationship showing a moderate relationship, whereas, deep lakes with lower colour presented a weaker relationship. Therefore, we assessed the use of CDOM spectral shape information to improve DOC estimates regardless of observed DOM differences due to climatic seasonality and lakes' morphometry. The use of well-known CDOM spectral shape metrics (i.e., S275-295 and a250:a365 ratio) significantly improved DOC estimation. Particularly, using a Gaussian decomposition approach we found that much of the variation in the spectral shape, associated with the variability of CDOM:DOC ratio, was explained by differences in two dynamic regions centred at 270 and 320 nm. A strong nonlinear relationship was found between the a270:a320 ratio and the DOC-specific absorption coefficients a*355 and a*440. This was translated into a further improvement in DOC estimation yielding the higher R2 and lower mean absolute differences (MAPD < 16%), either considering the entire data set or shallow and deep lakes separately. Our results highlight that incorporating the CDOM spectral shape information improves the characterization of the DOC pool of inland waters, which is particularly relevant for remote and/or inaccessible sites and has significant implications for the environmental management, biogeochemical studies and future remote sensing applications.
