The Effects of Surface Longwave Spectral Emissivity on Atmospheric Circulation and Convection over the Sahara and Sahel Regions

This study quantifies the impact of the inclusion of realistic surface spectral emissivity in the Sahara and Sahel regions on the simulated local climate and beyond. Spectral dependence of the surface emissivity has been ignored in current climate models, i.e. surface emissivity is assumed to be a constant for the entire longwave. But in reality the surface emissivity can vary with the spectral frequency. It can be as low as 0.6-0.7 over the infrared window band for deserts. This study incorporates realistic surface spectral emissivity in the Sahara and Sahel into the CESM version 1.1.1, while keeping treatments of surface emissivity for the rest of the globe unchanged. Both the modified and the standard CESM are then used to carry out a 13-year simulation with prescribed climatological sea surface temperatures and the outputs from the last 10 years are used for this study. The modified CESM simulates a warmer annual-mean surface air temperature by up to 1.2 K than the standard CESM in the Sahara and Sahel regions. The modified CESM also has a warmer and wetter troposphere than the standard CESM, with the largest differences seen in the planetary boundary layer. As a result, the modified CESM in general favors more convections in the Sahara and Sahel regions and leads to more convective rainfall, which is most noticeable in August with a 77% increase in the Sahara and a 19% increase in the Sahel with respect to the standard CESM. The impact of incorporating surface spectral emissivity in the Sahara and Sahel can be seen outside these regions as well. The simulated rainfall belt over West Africa in the modified CESM, compared to the standard CESM, increases in all seasons and reaches further poleward during its seasonal migration. These precipitation differences are associated with the differences in column moisture flux convergence. Besides, the modified CESM simulates a larger annual-mean rainfall at the coast region north to the Gulf of Guinea than the standard CESM does, which reduces the negative precipitation bias in the standard CESM with respect to the observational GPCP dataset.