Environmental & Climate Sciences Department Seminar

Ice clouds currently reflect ~17 W m-2 of shortwave radiation and trap ~22 W m-2 of longwave radiation on global average. However, if the distribution of cloud heights and microphysical properties changes in response to increases in greenhouse gases and aerosols, associated changes in the radiative impact of clouds could feed back on Earth's climate. Representations of ice particle density, scattering and sedimentation are needed for global and regional climate models that predict these effects. Parameterizations of other processes, such as riming, aggregation, sedimentation and evaporation, are also needed for numerical weather models that predict the destructive impact and quantitative precipitation forecasts for winter storms, hurricanes, mesoscale convective systems and other events. Further, algorithms retrieving cloud properties from ground- and satellite-based sensors require assumptions about ice crystal properties. To develop such parameterizations, accurate observations of ice particle sizes, shapes, phases and concentrations are needed. Techniques measuring these ice crystal properties are reviewed. Sources of uncertainty, related to statistical counting, variability in cloud properties for similar environmental conditions, and errors induced by the processing of data and the instruments themselves are discussed using data collected over Alaska, Australia, and the continental United States. It is shown that although there are still uncertainties in in-situ observations of small ice crystals due to potential shattering of large particles on probe tips and the limited resolution of state-of-the-art cloud particle imagers, progress on characterizing small crystals has been made. The use of instrumental and statistical uncertainties in the development of stochastic cloud parameterizations is then introduced. A specific application to the representation of mass-dimensional (m-D) relationships m=aDb is shown, where (a,b) are given as surf