Seasonal variability of precipitation over South Eastern South America: its relationship with upper level circulation as physical base for predictions
At the interannual timescale Sea Surface Temperature (SST) associated with El Niño Southern Oscillation (ENSO) have been identified as modulators of rainfall over South America. The proposed mechanism responsible for anomalies over northern South America involves the modification of the Walker circulation, while the generation (or excitement) of the Pacific-South American (PSA) modes would represent the teleconnection to the subtropics. However, the latter is present even in seasons in which ENSO is not at its peak. In this study our aim is to investigate the interannual variability of seasonal precipitation over South Eastern South America (SESA), which comprises southern Brazil, Uruguay, Paraguay and northeastern Argentina, and its dynamical dependence from ENSO and the PSA modes. We have shown the existence of a pervasive mode of upper level atmospheric variability which dominates the circulation over South America in all seasons. The mode consists of a continental scale vortex and resembles the eastern end of the leading Pacific-South American mode (PSA1). Differences in the structure and intensity of the vortex and the effects of orography affect the transport and convergence of moisture into SESA thus creating rainfall anomalies there. Motivated by and to test further the reported relationship between the upper level wind and precipitation anomalies, we developed a method to predict precipitation over SESA in which the upper level wind is the predictor. The method has a high potential in all seasons, but limited skill in forecast mode due to limitations of Coupled General Circulation Model performances. In exploring the predictability of the vortex, we assessed to what extent each connection of the schematic chain ENSO->PSA1->vortex->rainfall holds. The latter relationship between the vortex over SA and rainfall in SESA holds in spring, summer and fall. The vortex is, at least partially, an internal mode of variability in all seasons. In spring the whole chain of elements is observed: the vortex appears as a mode forced by ENSO via excitement of the leading PSA mode. In summer the vortex is uncoupled from the circulation over the Pacific Ocean, supporting the interpretation that regional effects dominate during the monsoon season. No connection with SSTs, and thus predictability, is found for this season. In fall the PSA1 pattern and the vortex are partially linked as for the spring season. However, the connection with SSTs is more puzzling and further analysis is required to clarify the nature of the leading PSA mode and its predictability. In view of our results, we underline that the leading Pacific-South American pattern properly comprehends centers of anomalies over the Southern Pacific Ocean only but not those over the South American sector.