A statistical study of seasonal winter rainfall prediction in the Comahue region (Argentina)

M.H GONZÁLEZ, N.M. SKANSI, F. LOSANO

Abstract

The aim of this study is to detect the possible causes of the May-to-July rainfall (MJJ) over the Limay and Neuquén river basins. In order to establish the existence of previous circulation patterns associated with interannual rainfall variability, the mean areal precipitation in each one of the basins was correlated to some predictors during the previous three month period (February-March-April). The result is that MJJ rainfall in both basins is related to sea surface temperature and geopotential heights at different levels previously observed in some specific areas of Indian and Pacific Oceans, probably due to wave trains which begin in those areas and then displace towards the Argentine Patagonia coast, thus generating precipitation systems. There are also observed significant correlations with sea surface temperature in the Atlantic Ocean over Brazil and the Argentine coast, associated with the water vapor income into the continent and with zonal and meridional wind over the basins, related to humid air advection. The prediction schemes, using multiple linear regressions, showed that the selected variables are the cause of the 51% of the MJJ rainfall variance in the Limay river basin and the 44% in the Neuquén river one. The scheme was validated by using a cross-validation method and significant correlations were detected between observed and forecast rainfall. The 2009 winter rainfall was analyzed and showed that circulation indicators were useful to predict winter rainfall.

Keywords

Seasonal prediction, precipitation, multiple linear regression, sea surface temperature, atmospheric circulation.

Full Text:

PDF