Tropical atlantic climate biases and DAMIP experiments: insights from CMIP6 models

This study investigates the representation of the Tropical Atlantic zonal and meridional modes of sea surface temperature variability, as well as precipitation in western Sahel, northeast Brazil, Amazon and equatorial Atlantic in a set 40 CMIP6 models under the historical period in both full-forcing historical simulations (hist-full) and the Detection and Attribution Model Intercomparison Project (DAMIP) single-forcing simulations: historical natural (hist-nat), historical aerosol (hist-aer) and historical greenhouse gases (hist-GHG). The mean state and variability represented in the CMIP6 historical full forcing simulations is compared to observations. The mean state and long term trends in the single forcing simulations is analysed and compared against the full-forcing and control simulations. Results show that there is a warm bias in the Atlantic Cold Tongue region observed in previous CMIP versions which persists in CMIP6. The models simulate higher mean precipitation rates than observations in northeastern Brazil and eastern equatorial Atlantic, while they present a negative bias overall in Amazon and Sahel regions. For the Detection and Attribution Model Intercomparison Project (DAMIP) experiments, aerosols lead to a decrease in climatological mean precipitation in the Amazon region while greenhouse gases lead to a warming trend in SSTs accompanied by a slight weakening of trade winds. Statistically significant long-term precipitation trends are observed only in Sahel region for the full-forcing simulations.