comparison between statistical precipitation prediction in northern patagonia (argentina) using era - interim and ncep reanalysis datasets

Abstract

Northern Patagonia is a dry area with scarce precipitation and agricultural and livestock activities which are highly dependent on inter-annual rainfall variability. In this study statistical regression models were derived to estimate summer precipitation in a single location using two different sets of data: ERA - INTERIM and NCEP reanalysis. Both datasets were used to determine the predictors. The models were derived using lineal backward regression techniques and the different efficiency between the derived models was analyzed. Results pointed out the best resolution of ERA-INTERIM dataset, allowed the definition of more and better predictors with physical meaning and improved efficiency.