Toward a Better Characterization of Aerosols, Clouds, and Aerosol-Cloud Interactions in Climate Modeling through Increasing Horizontal Resolution

Tuesday, 16 December 2014: 2:25 PM
Po-Lun Ma1, Philip J Rasch1, Minghuai Wang1, Hailong Wang1, Steven John Ghan1, Xiaohong Liu2, Richard C Easter1, William I Gustafson Jr1, Simone Tilmes3 and Hsi-Yen Ma4, (1)Pacific Northwest National Laboratory, Richland, WA, United States, (2)University of Wyoming, Laramie, WY, United States, (3)National Center for Atmospheric Research, Boulder, CO, United States, (4)Lawrence Livermore National Laboratory, Livermore, CA, United States
Aerosol-cloud interactions occur at sub-grid scale in most atmospheric models and remain as one major source of uncertainty in climate projection. The coarse (1-2 degrees) resolution typically employed in global climate models is insufficient to resolve these processes and is believed to lead to biases. In this study, we use the Community Atmosphere Model Version 5 (CAM5) running at grid spacing of 2, 1, 1/2, and 1/4 degrees, nudged toward the very high-resolution Year Of Tropical Convection (YOTC) analysis, to explore the resolution dependence of the model physics contributing to the simulation biases of aerosols, clouds, and aerosol-cloud interactions.

Using satellite retrievals as well as surface and aircraft measurements, we find that many biases commonly observed in global climate models are significantly reduced by increasing resolution. The high-resolution model produces a more realistic simulation of extreme events of aerosol concentration and precipitation, cloud susceptibility to aerosols, and aerosol transport into the Arctic. These characteristics are related to the change of balance of physics that leads to stronger accretion over autoconversion in warm clouds, stronger but less frequent cloud droplet nucleation, and less collocation of aerosols and clouds. The estimate of aerosol indirect forcing is greatly reduced as a result. Although many of the aerosol and cloud processes remain sub-grid scale even at the highest resolution we explore here, our results suggest that a better characterization of aerosols, clouds, and aerosol-cloud interactions can be achieved by increasing model resolution to better resolve meso-scale processes such as eddy transport and moisture convergence. Further improvement might rely on adequately addressing other sources of uncertainty in the model including the representation of physical processes (such as interactions between aerosols and convective clouds) and boundary forcing (such as aerosol emissions).