Identification of Homogeneous Precipitation Regions and Climate Projections with RegCM4 in the Andean Altiplano

Abstract

The climate of mountainous regions is not easy to study due to the difficulty in installing and maintaining in situ weather stations to record data. Knowledge of the climate of these areas is mainly obtained through remote sensing of the atmosphere and numerical modeling. In order to contribute to the knowledge of the Andean altiplano of Peru and Bolivia climate, this study aims to identify clusters with a similar annual precipitation cycle, as well as the projected climate changes for the altiplano. For that, it is used the regional climate model RegCM4 nested in three global climate models (GFDL, MPI e HadGEM2) of the Coupled Model Intercomparison Project Phase 5 (CMIP5) under RCP8.5 future scenario. The homogeneous clusters are defined using the precipitation from the meteorological stations of the DECADE project, and these data are also used to validate the simulations in the present climate (1981-2005). Cluster analysis divided the Andean altiplano in four clusters according to the characteristics of the annual cycle of precipitation. RegCM4 ensemble simulates the annual precipitation with a phase similar to the observed one but overestimates it in the four homogeneous clusters. For the future climate (2030-2060), the ensemble projects for these regions a trend of a slight increase in precipitation in summer (~5%) and a decrease in the winter, which may exceed 50% in July, as simulated by RegGFDL in cluster 4. Additionally, regional differences can be observed in the annual trend signal as a function of latitude and homogeneous clusters. Both factors indicate that there is variability in the strength of the signal projected by each member of the ensemble and, also, regional dependence.