Some atmospheric and oceanic indices as predictors of seasonal rainfall in the Del Plata Basin of Argentina

Different oceanic and near-global atmospheric indices are analyzed in connection to the variations of seasonal rainfall over northeastern Argentina in order to determine their potential as predictors. This region is part of the hydrological system of the Del Plata Basin, and is drained by the copious Parana and Uruguay rivers, two of the main courses flowing across the basin. Monthly precipitation, flow rate, Southern Oscillation Index (SOI), sea surface temperature anomaly (SSTs) Index for Nino 3 Region (N3) (1960-2003) and SSTs for different oceanic regions of the Southern Hemisphere (SH) Atlantic and Pacific Oceans (1981-2003) are used. A practical measure of rainfall deviations is obtained by means of the Standardized Precipitation Index (SPI). Significant correlations between SPI and near-global indices (based on SOI and SSTs) determined the months whose indices may be used as predictive variables. Using the Step-wise method, optimum predictors were chosen to generate the linear regression models. From a set of possible models, those with the best fitting statistical measures were considered and the ability of the proposed regression equations tested. The results for the study region showed that during SH spring, summer and early autumn, SPI values are related to SOI and N3, with individual monthly values, at 1- to 5-month lead times used as predictors. Besides, SPI links with SSTs of the Atlantic/Pacific Ocean along the whole year, especially during SH autumn and winter/SH late winter, spring and summer, with different month lags. It is worthwhile noting that almost year-round estimation of rainfall anomalies is possible by using the adequate predictive variable and forecast lead-time. (c) 2006 Elsevier B.V. All rights reserved.