Assessment of CMIP6 models' performance in simulating present-day climate in Brazil

• Autor(a) principal: Firpo, Mári Ândrea Feldman
• Data de Publicação: 2022
• Outros Autores: Guimarães, Bruno dos Santos, Dantas, Leydson Galvíncio, Silva, Marcelo Guatura Barbosa da, Alves, Lincoln Muniz, Chadwick, Robin, Llopart, Marta Pereira [UNESP], Oliveira, Gilvan Sampaio de
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.3389/fclim.2022.948499; http://hdl.handle.net/11449/246176
• topic: assessment; climate change; climate modeling; CMIP6; precipitation; South America; temperature

Brazil is one of the most vulnerable regions to extreme climate events, especially in recent decades, where these events posed a substantial threat to the socio-ecological system. This work underpins the provision of actionable information for society's response to climate variability and change. It provides a comprehensive assessment of the skill of the state-of-art Coupled Model Intercomparison Project, Phase 6 (CMIP6) models in simulating regional climate variability over Brazil during the present-day period. Different statistical analyses were employed to identify systematic biases and to choose the best subset of models to reduce uncertainties. The results show that models perform better for winter than summer precipitation, consistent with previous results in the literature. In both seasons, the worst performances were found for Northeast Brazil. Results also show that the models present deficiencies in simulating temperature over Amazonian regions. A good overall performance for precipitation and temperature in the La Plata Basin was found, in agreement with previous studies. Finally, the models with the highest ability in simulating monthly rainfall, aggregating all five Brazilian regions, were HadGEM3-GC31-MM, ACCESS-ESM1-5, IPSL-CM6A-LR, IPSL-CM6A-LR-INCA, and INM-CM4-8, while for monthly temperatures, they were CMCC-ESM2, CMCC-CM2-SR5, MRI-ESM2-0, BCC-ESM1, and HadGEM3-GC31-MM. The application of these results spans both past and possible future climates, supporting climate impact studies and providing information to climate policy and adaptation activities.