Improved Tropical Cirrus Parameterization for Global Climate Models
The goal of this research is to help improve climate prediction through better representation of the tropical microphysical cirrus properties in Global Climate Models (GCMs). Cirrus clouds are one of the biggest uncertainties in the radiative budget, crucial to the understanding of short- and long-term trends in climate. The detailed composition of the high altitude tropical cirrus clouds is not well known, and the study of tropical cirrus microphysics will help to improve how these clouds are represented in climate models.
Our previous tropical parameterization was based on 5,546 ice particles size distributions (SD) taken from all three case studies (3 flights) on April 1, April 4 and March 17 1993, during the Central Equatorial Pacific EXperiment (CEPEX). When that research was conducted, CEPEX was the only microphysics in-situ data sets of tropical cirrus. There are only a few tropical cirrus datasets. The Tropical Warm Pool International Cloud Experiment (TWP-ICE) in the area around Darwin, Australia in 2006 provides the most comprehensive available observations of tropical cirrus to date that include data from colder temperature range (less than -55°C). This research involves analysis of all available ice particles size distributions from the flights on January 27, January 29, and February 2 2006, which are the best cirrus microphysics data sets during 2006 TWP-ICE. The principal objective of this project is to provide a means of estimating bimodal size spectra for tropical ice clouds, which with analysis of the examined convective cloud systems (from their initial stages through to the decaying thin high level cirrus), and therefore with some informed assumptions about ice particle shape, will provide improved means of estimating ice cloud radiative properties in GCM.
We have developed a means of diagnosing ice particle size distributions (SD), as a function of temperature (T) and ice water content (IWC). This general approach is pursued here, but making greater use of the microphysical ice cloud measurements from a much more comprehensive data set. Recent work indicates that the thermal single scattering properties of ice clouds are not adequately determined by effective diameter (Deff ) and IWC alone, and that information on SD dispersion (e.g. degree of bimodality) is also required. The proposed parameterization satisfies this need, providing three gamma SD parameters for each mode of the bimodal distribution. Three of these six parameters are approximated as constant, while the other three are expressed in terms of cloud IWC and/or temperature. These simple inputs make the parameterization convenient for large scale models, such as GCMs. The proposed parameterization provides a physically rigorous means for parameterizing the mass sedimentation rates from ice clouds in a microphysical sense. Accurate knowledge of mass removal rates is critical for predicting cirrus IWCs and radiative properties. Thus, we use the TWP-ICE ARM In-Situ Data to parameterize the prognostic behavior of tropical ice mass sedimentation rates.
To date the representation of Deff, for the ice phase in GCMs has been highly uncertain, and is often treated as constant or nearly constant. GCM sensitivity tests using the Hadley Centre’s Unified Model indicate shortwave uncertainties of 30 W/m2. We will relate Deff to temperature, and will estimate Deff from TWP-ICE SD parameterization for different ice crystal shapes. Synoptic analysis techniques will be used to assess the origin of each sampled air mass at the cirrus level, in order to estimate likely ice crystal habit. Techniques available will include backward trajectory analysis on reanalysis datasets and/or other higher resolution numerically analyzed datasets.