Ice Cloud Size Distributions

Affiliation(s)PIProject periodFunded by
DAS Mitchell, David L 03/15/2006 - 03/14/2011 Department of Energy

Project Description

The proposed project is a collaboration between the Desert Research Institute (DRI) and University of Wisconsin's Cooperative Institute for Meteorological Satellite Studies (CIMSS). An outstanding problem that contributes considerable uncertainty to Global Climate Model (GCM) predictions of future climate is the characterization of ice particle sizes in cirrus clouds. Recent parameterizations of ice cloud effective particle size differ by a factor of three, which, for overcast conditions, often translate to changes in outgoing longwave radiation (OLR) of 55 W m-2 or more. Such differences not only play havoc with temperature prediction but also with precipitation, since associated differences in cirrus heating rates will manifest in the upper troposphere and may change the distribution of temperature and humidity with height. For example, a warmer upper troposphere will be less conducive to deep convection, thus affecting precipitation. It has now become apparent that the instruments aboard aircraft that measure the ice particle size distribution (PSD) in cirrus clouds often produce ice particle artifacts created by larger ice particles colliding and shattering upon the rim of the instrument inlet that samples ice particles. These ice fragment artifacts may be a key reason for the large uncertainty in effective particle size. One way to address this problem is to use remote sensing from the ground and space to evaluate cirrus particle size and the concentrations of these controversial small ice crystals. If successful, satellite remote sensing would provide global coverage and thus provide what is needed for GCMs. This project has developed methodologies in both satellite and ground-based remote sensing for retrieving estimates of the concentrations of small ice crystals (maximum dimension D < 60 microns). It is precisely these crystal sizes that are so difficult to measure, with the concentration of these crystals varying by 2-3 orders of magnitude, depending on the instrument used. The method takes advantage of an interesting property of light absorption-emission, sometimes referred to as photon tunneling. Tunneling is responsible for the absorption of radiation outside or beyond the physical cross-section of an ice crystal. Its contribution to absorption is greatest when (1) particle size and wavelength are comparable, (2) when the particle is quasi-spherical and (3) when the real index of refraction is high. Fortunately, condition (1) is satisfied only for small ice crystals at thermal infrared wavelengths and (2) is satisfied since small ice crystals tend to be quasi-spherical. Since the real index of refraction in ice abruptly increases between the 11 and 12 micron wavelengths in the infrared, the difference in heat emission between these two wavelengths is an ideal measure of the degree of photon tunneling in ice clouds and hence the concentration of small ice crystals. By retrieving the cloud emissivity at these two wavelengths along with cloud temperature, we can calculate the concentration ratio of small crystals to larger ice particles in a cirrus cloud, as well as the concentration of small ice crystals for a given cloud ice water content (IWC). Other cloud properties retrieved include the amount of ice in the cloud, known as the ice water path (IWP), and the effective diameter of the PSD. This knowledge will allow us to characterize the PSD of cirrus clouds in GCMs which is needed to calculate cirrus cloud radiative properties and their effect on climate. It will also provide estimates of IWP that are needed to test whether the predicted IWP in GCMs are realistic. By providing reasonable estimates of small ice crystal concentrations, it may provide insights relating these concentrations to natural and anthropogenic aerosol concentrations.