First optical observations in the turbidity maximum zone in the Río de la Plata estuary: a challenge for atmospheric correction algorithms

• Autores: Dogliotti, Ana Inés; Camiolo, Martina; Simionato, Claudia; Jaureguizar, Andrés Javier; Guerrero, Raúl Alberto; Lasta, Carlos Ángel
• Año de publicación: 2014
• Idioma: inglés
• Tipo de recurso: documento de conferencia
• Materia: Ecología; Oceanografía, Hidrología, Recursos Hídricos; Turbiedad del Agua; algoritmo de corrección atmosférica
• Institución: Comisión de Investigaciones Científicas de la Provincia de Buenos Aires

The Río de la Plata (RdP) estuary, located at 35°S on the southwestern Atlantic Ocean, is a shallow and large-scale plain, which drains the second largest basin in South America. The RdP river carries high amounts of nutrients, suspended particulate and dissolved organic matter to the adjacent shelf waters and is considered among the most turbid estuaries in the world. A turbidity maximum and a sharp surface front defining its seaward edge is a distinctive feature of this estuary. Such high sediment loads represent a challenge to atmospheric correction algorithms which usually rely on the assumption of zero waterleaving reflectance in the near infrared (NIR) or short wave infrared (SWIR) parts of the spectrum. Uncertainties of the primary remote sensing products have never been quantified in RdP before due to lack of in situ measurements. In February and April 2013 two field campaigns were performed in the turbidity maximum zone where water reflectance was measured and surface water samples were collected for turbidity and total suspended particle concentration determinations. A match-up analysis was performed to evaluate the performance of five atmospheric correction algorithms on MODIS-Aqua data that use the NIR and/or SWIR bands to estimate the aerosols optical properties in a pixel-by-pixel basis and from clear water pixels and then applied to the whole image. Satellite retrievals of remote sensing reflectance at the visible bands of Aqua sensor generally showed quite large uncertainties and constant underestimation (largest at the blue bands), whereas the uncertainties in the NIR bands were the lowest.