Influences of Surface Spectral Emissivity and Cloud Longwave Scattering on Climate Simulations

Abstract

Longwave (LW) radiation plays a critical role in the Earth’s climate system. It carries energy from the Earth to space, thereby balancing the global-averaged net solar radiation at the top of the atmosphere. It also redistributes energy within the atmosphere, and between the atmosphere and the surface. Due to its importance, a correct and faithful representation of the LW radiation processes in climate models is crucial for understanding the climate system and projecting future climate. Because comprehensive LW radiation calculations are computationally expensive, many approximations are made to accelerate these calculations. Two common approximations are blackbody surface and non-scattering clouds. At least twenty years ago, researchers argued the validity of these approximations, but these arguments have received relatively little attention until recently.

This dissertation, along with other recent studies, investigates the impact of surface spectral emissivity and ice cloud LW scattering on simulated climate. Specifically, this dissertation implements surface spectral emissivity, a two/four-stream LW radiative transfer solver, and a state-of-the-art ice cloud LW optical scheme into the Community Earth System Model version 1.1.1 (CESM1.1.1) and the DoE Energy Exascale Earth System Model (E3SM). Using the modified version of the CESM1.1.1, this dissertation investigates: (1) The surface emissivity effect over the Sahara and Sahel. The surface emissivity in these regions can be as low as 0.6-0.7 over the infrared window band while close to unity in other bands, but such spectral dependence has been ignored in climate models. The inclusion of realistic surface emissivity over the Sahara and Sahel, compared to the blackbody surface, increases the surface air temperature over these regions and produces more convective rainfall, especially in the Sahara. The precipitation south of the Sahel is also increased, indicating that the changes of surface emissivity can influence the local climate and beyond. (2) The ice cloud LW scattering effect on polar climate. Cloud LW scattering is usually neglected in climate models. The traditional rationale is that this scattering is negligible compared to strong LW absorption by clouds and greenhouse gases. This rationale, however, is not valid in the polar regions, in which the atmospheric absorption is weak due to the small amount of water vapor, implying that cloud LW scattering is not negligible anymore. Using CESM with a slab-ocean model, the scattering effect increases the Arctic (Antarctic) winter surface temperature by around 1.4K (1.4K). Interestingly, this effect becomes much weaker, only 0.1K (0.4K), when the sea surface temperatures and sea ice are prescribed. These results highlight the importance of the cloud LW scattering effect in the polar regions and the importance of surface-atmosphere coupling when this effect is considered. (3) The combined effect of surface emissivity and ice cloud LW scattering on polar climate. When a non-blackbody surface is combined with scattering clouds, multiple scattering between the surface and clouds can occur and retain additional energy in the Earth. The CESM simulations show that these two effects are linearly additive in the polar regions. This dissertation also shows that the modified E3SM, compared to the standard E3SM, reduces the prominent surface warm bias during the Arctic winter by half, mainly because of the new ice cloud optical scheme. The influences on other fields are minimal.

Altogether, this dissertation demonstrates the importance of surface spectral emissivity and cloud LW scattering on the simulated climate, particularly over the polar regions.