The Predictability of Drought and Pluvial Conditions over the Americas: The Role of the Oceans and Land

In this study we examine the nature and predictability of extended periods of wet and dry conditions over the Americas using seasonal hindcasts as well as century-long simulations with the NASA Seasonal-to- Interannual Prediction Project (NSIPP-1 atmospheric general circulation model forced with observed SSTs. Moderately large ensembles of runs allow us to diagnose the changes over time in predictability assuming the SSTs are known. The regional and temporal variations in enhanced and reduced predictability in both drought and pluvial conditions are diagnosed in terms of changes in the "signal" (forced by the SST) and "noise" (determined by the unpredictable components of atmospheric and land variability). The variations in predictability are interpreted in terms of their dependence on the strength of soil moisture feedbacks, long-term changes in the SST, as well as ENSO.