The Icy Business of Calculating Cirrus Clouds
Somerville, R. C., Scripps Institution of Oceanography
General Circulation and Single Column Models/Parameterizations
Cloud Modeling
McFarquhar, G.M., S. Iacobellis, and R.C.J. Somerville, 2003: "SCM simulations of tropical ice clouds using observationally based parameterizations of microphysics," J. Climate 16(11):1643-1664.
Iacobellis, S.F., G.M. McFarquhar, D.L. Mitchell, and R.C.J. Somerville, 2003: "The sensitivity of radiative fluxes to parameterized cloud microphysics," J. Climate 16(18): 2979-2996.
Three-part graph: The use of a new cloud parameterization in the Scripps Single-Column Model, in red, demonstrates how the new information allows greater agreement with actual data, in black, from the same 18-month time period at ARM sites in the Southern Great Plains, the Tropical Western Pacific and the North Slope of Alaska. The measurements were of downwelling solar radiation, which is the direct energy from the sun plus the diffuse energy from scattered sunlight, over an 18-month period. Note that the conventional numerical weather model, shown in blue, predicted less radiation than the measurements indicate in the tropics and more radiation in the Arctic. The single-column model, when run with the new cloud parameterizations is in closer agreement to the measurements at all three sites.
Key Participants: Sam Iacobellis, Scripps Institute of Oceanography
Richard Somerville, Scripps Institute of Oceanography
Greg McFarquhar, University of Illinois at Urbana-Champaign
David Mitchell, Desert Research Institute
Clouds and their feedbacks pose the largest single source of uncertainty in climate models. Fresh insights about how clouds contribute to the Earth's ability to reflect away sunlight and contribute to the greenhouse effect were translated into a new parameterization (a formula that makes models run) that greatly improves model predictions. When compared to long-term data collected from the Atmospheric Radiation Measurement Program sites in the Southern Great Plains, the Tropical Western Pacific and the North Slope of Alaska, results using the new parameterization were nearly identical to measurements. The achievement will allow climate modelers to confront the most advanced theories with high-quality measurements of how real clouds behave, according to Richard Somerville, one of the members of the investigative team. They will likely lead to significant improvements in our understanding of climate and our ability to simulate it.
The advance started with development of theories to account for the physical properties of ice crystals that make up cirrus clouds, which occur roughly 6 to 12 miles above the Earth's surface. These particles exist in a dizzying array of sizes and shapes, and the amount of ice in a given volume of air also varies greatly from one cloud to another. The physical details of the crystal determine how each particle reflects light from the sun and absorbs infrared heat from the Earth as these forms of energy pass through the atmosphere.
After key participants included Greg McFarquhar and David Mitchell defined the physical properties of cirrus ice crystals, relationships were calculated between meteorological factors, such as temperature and water amount, and the ability of cirrus clouds to reflect and absorb energy. Somerville and his colleague Sam Iacobellis then incorporated these relationships in the form of parameterizations within their computer model that is used to simulate the volume of atmosphere over the three ARM sites.
In their model, the two scientists included treatments of all the physical processes typically included in the most advanced global climate simulations, including feedbacks between clouds, sunlight and infrared energy, and the water cycle.
Actual measurements from the ARM sites were used to assess the realism of the model simulations, which were carried out with both conventional treatments of cirrus cloud ice particles and the new ice crystal theories. The model results showed excellent agreement with ground-based and satellite measurements using the new theories. Additionally, a numerical weather forecasting model was run for the same time period. No enhancements were made to this model.
The scientists discovered that the single-column model simulation of broad-scale features was highly sensitive to how they incorporated the influence of cloud ice crystals. Measurements from meteorological data at the ARM sites showed the model results underestimated the average size of ice crystal sizes found at the highest altitudes as well as the range of sizes. Somehow the clouds contained larger crystals on average and a much wider variety of sizes than the model could explain. They theorized that one possible cause of the discrepancy might lay in an underestimation by the model of the total amount of water contained in the clouds in the form of ice, rather than a fundamental inadequacy in the theory itself.
By varying the size of the ice crystals and the range of sizes at a given temperature and total water amount, they calculated an estimate of the effects of these theoretical shortcomings on the computed rates at which the atmosphere cools by emitting infrared energy to space as the cloud particles modified the greenhouse effect, and also on the ability of the clouds to reflect sunlight. These effects were substantial, amounting to as much as 20 percent when averaged over a season, and even larger effects were observed in a single day. Their analysis showed the sensitivity of models to the way cloud effects are incorporated.
Armed with new information about the microphysical aspects of cloud ice crystals, they can begin to work on more advanced theories of cloud behavior. As a next step, the research team plans to incorporate the new parameterizations into major global weather prediction models to see what they do as weather forecasts.
The U.S. Department of Energy's Atmospheric Radiation Measurement Program supported this study. Leading the study were Sam Iacobellis and Richard Somerville, Scripps Institution of Oceanography. Other key participants included Greg McFarquhar, University of Illinois at Urbana-Champaign, and David Mitchell, Desert Research Institute.
