CONICET | Buscador de Institutos y Recursos Humanos

This paper attempts to understand the rainfall interannual variability in order to predict seasonal precipitation using a linear regression model. This work analyze summer precipitation (December, January and February) in central-west Argentina based on comparing the behavior of previous (in November) atmospheric and oceanic variables in extreme years with drought and excess precipitation. The main predictors were geopotential height and warming of central Pacific Ocean. Generally,higher than normal precipitation cases were associated with the weakening of the subtropical high and subpolar low, a weakening of zonal flow and therefore greater energy exchange between low and high latitudes and warm anomalies of sea surface temperature in November in the eastern equatorial Pacific. This analysis allowed us to define, predictors that were used to generate a multiple linear regression model using forward stepwise methodology which explained 28% of variance of summer precipitation.