Análisis comparativo de Downscaling estadístico y dinámico en las cuencas de los ríos Paute y Jubones

Abstract

Global circulation models are a powerful tool for climate prediction. The coarse scale of their results makes them difficult to apply to decision-making processes at local and regional level. Aiming at the incorporation of regional and local information, several downscaling techniques have been devised. Notwithstanding, their application, the results exhibit errors; to tackle this problem, the technique known as Quantile Mapping allows a consistent change on the results distribution, so that they are fitted to the observations distribution. This technique is applied to downscaling results for the Paute and Jubones basins, located in Southern Ecuador. It is observed that they depend on the Quantile Mapping variant applied. Even though it is clear that errors cannot be completely eliminated, improvement of up to 70% has been attained in dynamical downscaling; in statistical downscaling, just a slight improvement is observed. To identify and generate projections of climate change in the basins, Delta method was applied, with their results seasonality, anomalies, and climate variability were analyzed.