Isosurface of cloud water depicting the outer boundary of a trade cumulus cloud in simulations for studying the effects of aerosol pollution on clouds and hence global climate. The colored contours denote the liquid water content where the cloud intersects the edge of the model domain

Owen Toon and Peter Colarco, Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder
Andrew Ackerman, NASA Ames Research Center

Research Objectives
The technical objective is to build the most advanced explicit aerosol and cloud microphysics model. The scientific objectives are (1) to explore the factors that determine cloud fraction in tradewind cumulus, (2) to use the explicit microphysical model to evaluate the precipitation parameterizations used in simpler models, (3) to investigate the relationship between cloud optical thickness and sea surface temperatures, and (4) to investigate the evolution of the aerosol size distribution and its effect on the atmosphere's radiative budget.

Computational Approach
For the dynamics, we solve the anelastic form of the Navier-Stokes equations in flux form on a rectangular grid using finite difference techniques. The advection algorithm is second-order accurate for smooth flows. Sub-grid scale mixing is solved through either a first-order closure or a turbulent kinetic energy model. For the aerosol and cloud microphysics code, the particle size distributions are resolved on a fixed mass grid with size bins of geometrically increasing width. Condensational growth is treated using the piecewise polynomial method for advection in mass space. For the plane-parallel radiative transfer model, particle scattering and absorption coefficients are calculated using Mie theory. Gaseous absorption and emission are treated using an exponential sum formulation. For the mineral dust simulations, we incorporated the NCAR Model for Atmospheric Transport and Chemistry (MATCH) into the aerosol code.

Accomplishments
In our simulations of tradewind cumulus, we found that cloud coverage depends strongly on the sub-grid scale (sgs) mixing. Previously we assumed the sgs mixing length scale was fixed by the model resolution, but we obtained realistic results when the sgs mixing length decreased with increasing atmospheric stability if precipitation was included.

In February and March 1999, a dark, murky haze was observed throughout the Indian Ocean, and there were very few clouds. We used our large eddy simulation model to investigate the possibility that solar heating due to the carbonaceous haze evaporated the clouds. Our results indicate that a heating rate at noon of only 1° per day is enough to significantly reduce the cloud coverage, contradicting the widespread assumption that increased aerosol loadings result in more clouds that reflect solar energy back into space.

We also applied the model to a case study of desert dust transport during the ACE-2 experiment. These simulations showed good agreement with observed vertical profiles of dust optical depth and particle sizes, and captured the general radiative properties of the aerosol layer.

Significance
The indirect effect of aerosols on the radiative heat budget (through their impact on cloud properties) represents a leading uncertainty in the effects of mankind on the global climate. On the coarse grids of general circulation models, the representations of clouds (and their interactions with aerosols) are necessarily crude. Our work aims to advance understanding of cloud processes so that they may be more realistically treated in large-scale models.

Publications
A. S. Ackerman, O. B. Toon, D. E. Stevens, A. J. Heymsfield, V. Ramanathan, and E. J. Welton, "Reduction of tropical cloudiness by soot," Science 288, 1042 (2000).

P. R. Colarco and O. B. Toon, "Three dimensional model simulations of the desert aerosol lifecycle: Comparisons to ground and satellite observations during ACE2-2," ICTP Summer Conference on Chemistry-Climate Interactions, Trieste, Italy, June 2000.

J. P. Taylor and A. S. Ackerman, "A case study of pronounced perturbations to cloud properties and boundary layer dynamics due to aerosol emissions," Q. J. Roy. Met. Soc. 125, 2643 (1999).