Resolution dependence of CO2-induced Tropical Atlantic sector climate changes

A long-standing problem in state-of-the-art climate models is the Tropical Atlantic (TA) warm sea surface temperature (SST) bias, which goes along with major biases in large-scale atmospheric circulation. Here we show that TA-sector climate changes forced by increasing atmospheric carbon dioxide (CO2) levels are sensitive to model resolution. Two versions of a climate model employing greatly varying atmospheric resolution and exhibiting very different warm bias strength are compared. The version with high atmospheric resolution features a small SST bias and simulates an eastward amplified SST warming over the equatorial Atlantic, in line with the observed SST trends since the mid-20th century. On the contrary, the version with coarse atmospheric resolution exhibits a large SST bias and projects relatively uniform SST changes across the equatorial Atlantic. In both model versions, the warming pattern resembles the pattern of interannual SST variability simulated under present-day conditions. Atmospheric changes also vastly differ among the two climate model versions. In the version with small SST bias, a deep atmospheric response is simulated with a major change in the Walker circulation and strongly enhanced rainfall over the equatorial region, whereas the atmospheric response is much weaker and of rather different character in the model with large SST bias. This study suggests that higher atmospheric resolution in climate models may enhance global warming projections over the TA sector.