Downscaling of climate extremes over South America – Part I: Model evaluation in the reference climate

In this paper, we evaluate temperature and precipitation trends over South America (SA), simulated by the regional Eta model with 20-km horizontal resolution nested to three global models: The Canadian Earth System Model Second Generation (CanESM2-ES), The Met Office Hadley Centre Global Environmental Model - Earth System (HadGEM2-ES) and the Model for Interdisciplinary Research on Climate version 5 (MIROC5). The simulated trends are compared to observed trends (OBS) using six (four) extreme temperature (precipitation) indices recommended by the Expert Team on Climate Change Detection and Indices. In both simulations and OBS, extreme temperature trends over Amazon (AMZ) and Northeast Brazil (NEB) are higher than over South East SA (SESA) and West SA (WSA). Also, trends based on maximum temperatures are, in general, smaller than those based on minimum temperatures. The three downscaling simulations reproduce these features. The temperature simulations of the Eta-CanESM2 (Eta-MIROC5) show the largest (smallest) trends among the three simulations. The Eta-MIROC5, performs better than the other two simulations, reproducing the signals and the magnitudes of the extreme temperature trends, and some cooling trends over SESA, as shown in the observation. Precipitation trends from the three simulations do not compare well with observations, such as temperature trends. However, Eta-MIROC5 presents the best simulations, showing increase in the magnitude of the precipitation extremes over most of SA. The results from the three simulations presented here will be further used for the assessment of climate change projections in the SA region.