Vol. 10, No. 2, 119-130, 2011

Abstract
The La Plata River, located at 35°S on the Atlantic coast, is one of the largest waterways of South America. It carries a large amount of suspended particulate and dissolved organic matter, and is considered among the most turbid waters of the world. Very high values of total suspended matter have been reported in this region, with mean values ranging from 100 to 300 g m^-3 and extreme concentrations up to 400 g m^-3. Satellite sensors have shown to be the best tools available to map river plumes and to study their influence on the adjacent ocean. However, global algorithms for remotely estimating sediment concentration are not currently available. Moreover, such high sediment loads represent a challenge to atmospheric correction algorithms which usually rely on the assumption of zero water-leaving reflectance in the near infrared or short wave infrared part of the spectrum (black pixel assumption). In the extremely turbid waters of La Plata Estuary such assumptions are not valid. A two band algorithm to estimate turbidity using near infrared and the short wave infrared bands (858 nm and 1240 nm) of the MODIS-Aqua sensor is presented. The model is calibrated using in situ reflectance and turbidity measurements from turbid waters of the Southern North Sea and Scheldt River (Belgium) and then applied to MODIS imagery of La Plata River estuary (Argentina). A good correlation was found between modelled and in situ turbidity values when the algorithm was applied to concurrent MODIS imagery. Moreover, satellite-derived turbidity maps show a spatial distribution of sediment consistent with patterns and characteristic features of the estuary.

ISSN 1729-3782