The influence of diffetent wind stress forcing on Braszil Current - Eddy - Upwelling System off Cabo Frio (23°S)

Serrato, G.M.S. (1,2); Calado, L.(1); Soutelino, R.G.(1,3).

1) IEAPM - Brazilian Navy
2) CEM/UFPR - Federal University of Parana
3) IO/USP - University of São Paulo

The coastal upwelling at Cabo Frio (CF, southeast coast of Brazil) and neighboring regions is a well known seasonal process and mainly forced by the prevailing northeasterly winds. During each upwelling event, the wind spatial scales and variability in the region plus the influence of the Brazil Current’s (BC) mesoscale activity can lead to different pathways of the upwelling plume. This may determine the amount of upwelled water on the continental shelf surface during such events. Our study focuses on the interaction between regional winds and dynamical processes near CF. Several comparisons are carried out between observed (satellite) and modeled SST. The modeled SST fields are obtained from 12-day hindcasts using three different atmospheric forcings. The goal is to determinate which forcing dataset best predicts the upwelling features in order to build a reliable operational forecasting system for this area.

This upwelling forecast system is initialized by the Feature Oriented Regional Model System (FORMS) technique, which combines T-S climatology, cloud-free high resolution SST (GHRSST), and a BC parametric feature model to create a realistic nowcast field. Results from four ROMS experiments will be presented. The initial conditions are always the same, but different atmospheric forcings are used: i) NCEP/reanalisys I, ii) Global Forecast System (GFS), iii) MASTER/USP Atmospheric Forecast, and iv) no forcing (control run). All the wind-forced simulations exhibit a high correlation between the average modeled SST and satellite observations. After 12 days, the modeled upwelling plume shows similar distribution to the satellite observations in the three forcing scenarios. The simulation forced by the MASTER product had the lowest RMS and the closest upwelling plume area. The no-forcing experiment showed less SST variability in the upwelling plume, which is in agreement with other studies that have reported winds as the main forcing mechanism. It seems that the FORMS initialization technique works very well in this case. All three forced simulations reproduce the temporal plume variability well, with small differences.