Seasonal Sea Level Anomaly Patterns Over Argentine Continental Shelf

LAURA RUIZ-ETCHEVERRY; MARTIN SARACENO

Simposio; 20 years of Progress in Radar Altimetry symposium; 2012

Argentine Continental shelf circulation has been studied principally with numerical models. Results of these models have shown a mean northeastward flow south of 40°S, e.g. Palma et. al. (2008). The seasonal pattern has been described also with in-situ data but only for a limited number of dates corresponding to oceanographic campaigns. Satellite data provide an opportunity to solve the spatial and temporal issue. In particular, the 20-year database of sea surface height collected by radar altimeters is a unique dataset to study the circulation in regions with low density of in-situ data. Altimetry over continental shelf and coastal regions has improved considerably in the last few years, allowing researchers to study these complex areas. The aim of this work is to study the shelf circulation with gridded altimetry data produced by AVISO Archiving, Validation and Interpretation of Satellite Oceanographic data) and along-track altimetry data produced by CTOH (Centre de Topographie des Oceans et de l?Hydrosphere) over the Argentine Continental Shelf (30ºS-60ºS). Our first objective was to estimate the accuracy of the two data sets with in-situ data. Results show that error decreases (from 8 cm to 5.6 cm) for periods larger than 20 days. We then analyse the seasonal component of 18 years of SLA (sea level anomaly). The first 3 modes of variability explain more than 90% of the variance. Decomposition in Empirical Orthogonal Functions (EOF) of the seasonal cycle suggests that the first mode is related to the radiative cycle. The spatial pattern shows a meridional gradient between 35°S and 45°S and a zonal gradient north of Montevideo. These features cannot be explained by the radiative cycle. The second and the third mode explain only the 8.20 % and 1.34 % respectively, but do show a well-defined pattern which are discussed as a function of other variables.