Model & Observations

To assess the realism of the seasonal variations simulated by the model we compare the summer (DJF) and winter (JJA) SSS fields from the model with in-situ and Aquarius observations. The in-situ observations include 34090 bottle and CTD observations collected between 1911 and 2004. Data gaps of the historic observations, all of them in the deep ocean region, were filled with data taken from the World Ocean Atlas. To reduce the biases introduced by land contamination we masked Aquarius data closer than 100 km to the coast. Model and in-situ observations show similar seasonal variations of the SSS over the shelf. Both data sets show that, during the austral winter, there is a narrow tongue of freshwater extending northward along the South American coast, while during the austral summer this tongue retracts southward and expands offshore. The summer retraction of the Rio de la Plata waters creates positive SSS seasonal anomalies (annual mean minus the seasonal values) in the northern portion of the shelf—particularly along the boundaries of Uruguay and southern Brazil—and negative SSS anomalies in the southern domain. This situation is reversed during the winter months when, under the influence of downwelling winds, the plume becomes trapped to the coast and extends northward up to approximately 28˚S. This northward displacement generates large negative anomalies in the downstream region and positive anomalies in the upstream region. Aquarius cannot detect the nearshore displacement of the low salinity plume, due to the land mask. Aquarius, however, also shows an offshore expansion of the low salinity waters during the summer months that are consistent with the model and the in-situ observations.

 

The Seasonal Cycle

 

Matano et al. (2014) J. Geophys. Res. Oc. Submitted.

Although the largest SSS variations are observed in the inner and the middle shelf, the three datasets show consistent SSS signatures associated with the detrainment of the shelf waters into the deep ocean. During the austral summer, for example, the in-situ observations show a C-shaped, low-salinity tongue extending from the shelf into the deep ocean south of 34°S. The model reproduces a similar structure, albeit with smaller SSS gradients. The C-shaped tongue is not well defined in the Aquarius data, due to the lack of spatial resolution. Aquarius data, however, also show a low salinity intrusion extending from the continental shelfbreak into the deep ocean. Considering the different origins and the different time spans encompassed by our three sources of information, the similarity of the SSS patterns is quite reasonable. As shown in Section 5 the intrusion of the shelf waters into the deep ocean is largely controlled by the dynamics of the western boundary current. In fact, the C-shaped tongued just described roughly follows the mean location of the BMC.