Improved Simulation of Boundary Layer Clouds

Ghan, S. J., Pacific Northwest National Laboratory

Atmospheric Thermodynamics and Vertical Structures

Cloud Properties


Figure 1. Comparison of Boundary Layer Clouds Schemes in Climate Models with Satellite Observations

Figure 1. Comparison of Boundary Layer Clouds Schemes in Climate Models with Satellite Observations

Key Contributors: James McCaa, as part of his Ph.D. dissertation at University of Washington
Chris Bretherton, University of Washington
Dennis Hartmann, University of Washington
Steven Ghan, Pacific Northwest National Laboratory

Marine boundary layer clouds are among the most difficult clouds to represent in climate models. A team of atmospheric scientists from the University of Washington (UW) and the Pacific Northwest National Laboratory (PNNL) has developed a new physically-based parameterization of shallow convection that greatly improves the simulation of boundary layer clouds and their influence on the global energy balance. This advance has already been successfully tested against observations and is expected to substantially improve climate simulations by coupled atmosphere-ocean circulation models, leading to more realistic estimates of global climate change.

Marine boundary layer clouds arise from a complex interaction between turbulence, radiative heating, cloud microphysics, and moist convection, all occurring on spatial scales too small to be explicitly resolved by climate models. Climate models have either underestimated marine boundary layer clouds by inadequately treating the cloud-top radiative cooling that drives the clouds over cool ocean surface, or overestimated the clouds by inadequately treating cloud top entrainment and shallow moist convection, processes that ventilate moisture from the moist boundary layer to the free troposphere. Figure 1 shows how a poor treatment of boundary layer clouds (the HIRPBL scheme in the figure, HIRPBL is a common turbulence scheme) in a climate model causes large errors in the upwelling solar radiative flux at the top of the atmosphere. Excessive boundary layer cloud produces excessive upwelling radiative flux. Replacing the HIRPBL scheme with a scheme developed at the University of Washington (the UW scheme in the figure) improves the simulated radiative flux, but too much boundary layer cloud is still simulated across most of the Northeast Pacific. However, with the addition of a new physically-based parameterization of shallow convection (the UW+ShCu scheme in the figure) the simulated radiative flux agrees with the satellite observations remarkably well.

The shallow convection parameterization has been developed using a regional atmospheric circulation model as a test-bed. The next step will be to apply it to a global atmospheric circulation model, and finally to a coupled atmosphere-ocean circulation model. Once validated, the parameterization can be used to improve coupled atmosphere-ocean simulations of global climate change.

This study was supported by the ARM Program. The shallow convection parameterization was developed by James McCaa as part of his Ph.D. dissertation at UW. Chris Bretherton and Dennis Hartmann of the UW and Steven Ghan of PNNL were contributors and served as McCaa's dissertation advisors.