Abstract

Precipitation forecast is fundamental for improving the management of water resources, for development projects and risk reduction. Due to its high variability, high quality rain forecasts are still a challenge. In the present study, a weather generator (WG) was used to study the quality of the forecast of daily rainfall occurrence. The WG was implemented using three variables as predictors: the binary variable precipitation occurrence of the previous day (Kt-1), the maximum and minimum temperatures of the previous day (Txt-1, Tnt-1, respectively); and two co-variables: monthly values of the observed Sea Surface Temperature Anomalies of the Regions 1+2 and 3.4 of El Niño-Southern Oscillation (Niño 1+2 and Niño 3.4 respectively). We found that the variables Kt-1, Tnt-1 and co-variable Niño 3.4 are those that improve the performance of the precipitation occurrence forecasting. There was a noticeable difference in the number of consecutive wet and dry spell days in the altitudinal gradient in a rainy period. We propose several hypotheses based on the use of WG, which allow the understanding of the functioning of the climate system and the improvement of the forecast of precipitation occurrence in a mountainous area.