In the high latitudes during cold seasons, a substantial portion of precipitation falls in the form of snow. Measuring snow precipitation has many applications such as forecasting hazardous weather, understanding hydrological water budget and evaluating the cooling and freshening effects of snow onto ocean surface. However, unlike rainfall, snowfall measurement is extremely limited due to technical difficulties. Over ocean and in the uninhabited polar regions, perhaps there is no snowfall observation of any kind. The CloudSat Cloud Profiling Radar (CPR) is the first spaceborne radar orbiting the high latitudes and has the sensitivity to detect snowfall. The CloudSat mission thus provides a great opportunity to measure snowfall from space. First, the radar reflectivity can be used to retrieve snowfall. Second, the snow profile statistics derived from CPR can be used to build a snow cloud database that can be used to develop snowfall retrieval algorithms for other remote sensors, such as high frequency microwave radiometers available on current and future satellites. Given the fact that CPR only observes a ~1.5 km width strip with each orbit, the second application seems to be more important for global snowfall measurement. We have been developing a snowfall retrieval algorithm using high frequency (>85 GHz) microwave data from aircraft and satellite for the past several years. Based on the experience gained from the above exercise, we propose to develop a snow cloud database using CloudSat and other A-train data over a global (excluding areas where it never snows) scale. This database will benefit all precipitation retrieval and modeling community.
The proposed research consists of the following components: (1) Develop a radar reflectivity (Z) - snowfall rate (S) relation for CloudSat CPR based on backscattering of nonspherical snowflakes calculated using DDA modeling. (2) Derive snowfall rate profiles from the Z-S relation. By combing AMSR-E retrievals and numerical model analysis, we build a database of snow cloud profiles with snow and liquid cloud layers. (3) Perform radiative transfer modeling at high microwave frequencies, in which the nonsphericity of snowflakes is accounted. The snow cloud profiles and corresponding brightness temperatures constitute a priori database for further snowfall retrieval algorithm development using available satellite sensors, such as HSB (AQUA), SSMIS, and AMSU-B. It is noted that snowfall rate is not a standard product of CloudSat.
Through this project, we provide not only CPR-based snowfall retrievals, but also an observational-data-based database that can be used for developing snowfall retrieval algorithms. The PI has been working in the area of satellite remote sensing of cloud and precipitation for about 20 years. If selected, he is willing to serve as a member of the CloudSat science team.
