El Nino and prediction of anomalous monthly rainfalls in Ecuador

The coastal regions of Northern Peru and Southern Ecuador are well known for their sensitivity to the El Nino/Southern Oscillation (ENSO) phenomenon. Abundant rainfall is generally related to hot weather in the eastern tropical Pacific and to the weakened South Pacific trade winds, which allow the inter-tropical convergence zone (ITCZ) to migrate far to the south and to convey warm moist air to the coastal area. A new monthly rainfall index series has been established using data from a network of 75 rainfall stations in Ecuador's coastal region. Throughout the analysis, a distinction is made between a 'dependent' data set (1964-1982) used as a training period and an 'independent' portion of the record (1983-1993) reserved for validation. Multiple regression models are developed to predict monthly rainfall in the Guayaquil area, using as predictors precipitation, sea surface temperature (SST), meridional and zonal wind in the eastern equatorial Pacific, with all predictions being based on a one-month lead time. The resulting equations were used to predict rainfall anomalies in the independent data set. There is considerable predictive power for the rainy months of the year, the best predictability being for the period from March to May. The Multiple linear correlations explain 60-82% of the monthly precipitation variance. The north region's pre-season rainfall is the most powerful predictor for the rainy season peak in Guayaquil, whereas the eastern equatorial Pacific SST is the most powerful predictor for the end of the rainy season. Copyright (C) 2009 John Wiley & Sons, Ltd.