Regional Differences in South American Monsoon Onset: Implications for Onset Predictability

An objective of NOAA's Climate Prediction Program for the Americas (CPPA) is to characterize the predictability and improve the predictive capability of the South American Monsoon System (SAMS) onset. The main goals of this CPPA-funded project are to examine the mechanisms for SAMS onset and to assess its predictability. The central hypothesis is that detailed observations of regional differences in the structure and organization of convection during the SAMS onset will improve our understanding of key onset mechanisms, which in turn will allow us to better assess the predictability of SAMS onset using a GCM. An important theme is that regional differences in onset mechanisms have important ramifications for onset predictability. Previous studies suggest that the structure and intensity of convection in western and eastern Amazonia are different. We will build on that result to identify large-scale controls on onset, and test the hypothesis that the monsoon in eastern tropical South America (controlled by slowly varying Atlantic SSTs) is more predictable than in western tropical South America (controlled by more stochastic baroclinic systems). Our approach is to integrate observations and results from GCM simulations. These include observations of rainfall, convective organization and vertical structure from TRMM satellite data, large-scale circulation from NCEP reanalysis data, and assessment of onset predictability in an ensemble of NSIPP tier-2 GCM simulations.

In this poster we present early results from the observational portion of the project. Using the Global Precipitation Climatology Project (GPCP) dataset, wet season onset dates are determined across tropical and subtropical South America for each year between 1979 and 2006 in order to characterize regional differences in onset date for that period. In western Amazonia, where cold front intrusions associated with baroclinic waves have been shown to play a role in monsoon onset there, cold front episodes are identified using 24-hour sea level pressure tendency in NCEP reanalysis data. These results provide the context for later analysis of regional differences convective structure and organization from the TRMM satellite. In this way we will build a clearer picture of large-scale onset mechanisms, and set the stage for predictability studies.