Future Projections of Extreme Precipitation Climate Indices over South America Based on CORDEX-CORE Multimodel Ensemble

Abstract

Climate change is undeniable, and its effects, like droughts and intense precipitation, evidence this reality. However, many questions remain unanswered, such as climate extremes’ future frequency and intensity. International collaboration projects such as the Coordinated Regional Climate Downscaling Experiment (CORDEX)-Coordinated Output for Regional Evaluations (CORE) have emerged to address these questions, creating protocols to facilitate the intercomparison among simulations obtained by different regional climate models (RCMs). Given this context, this study aims to describe the projected changes in the extreme precipitation events over the South America domain by using a set of projections of three RCMs (Eta, RegCM, and REMO) under the CORDEX-CORE framework (except Eta). These models were nested in different global climate models (GCMs) from Coupled Model Intercomparison Project phase 5 (CMIP5) under the RCP8.5 scenario and integrated with a grid space of 25 km. In addition, six climate indices (PRCPTOT, SDII, R95p, P95, CDD, and CWD) were computed by season for the historical period (1995–2014) and four future time-slices (2020–2039, 2040–2059, 2060–2079, and 2080–2099). For brevity, this study focuses on the austral summer. RCMs are generally capable of representing the spatial pattern of climate indices but with differences in intensity. The ensemble with all RCMs shows a slightly better spatial correlation with the reference data than the ensemble of each RCM. The signal of the projections is similar between RegCM and REMO and, in general, opposite to Eta (mainly in the South Atlantic Convergence Zone region). Of the six climate indices analyzed, the ensemble performed with the three RCMs projects statistically significant negative (positive) trends for four indices in the Amazon (La Plata Basin), indicating a signal of dry (wet) conditions increasing towards the end of the century. Furthermore, both regions have significant positive (negative) trends of consecutive dry (wet) days. The region with higher uncertainties lies over southeastern Brazil, where Eta projects a dry climate and RegCM and REMO, a wet one.