Falling snow plays an important role in hydrological processes and climatic cycles of the Earth. The objective of this proposal is to investigate the challenges of falling snow detection and estimation algorithms for over land surfaces and to prepare for the next generation of such operational algorithms. Two of the major challenges addressed in the proposed work include (1) identifying and mitigating the effects of land surface signatures from brightness temperature observations, and (2) adequately modeling the physical characteristics of precipitating snow in radiative transfer calculations. Our investigations will ensure a physical consistency between the underlying cloud and atmospheric state and the observations, including levels down to the relationships between the single scattering radiative properties, microphysical shapes, and ice-air-water mixtures of a snowflake. We have a plan and methodologies to address variations in land surface emissivity and temperature and to determine how these affect the retrievals. The algorithm will employ wide band observations (e.g., at all Global Precipitation Measurement (GPM) Microwave Imager-(GMI) channels, 10-183 GHz) because cloud profiles have a continuum of states between all liquid (warm rain events) to partially liquid and frozen (with a melting layer) to all frozen (falling snow). The use of all channels will help distinguish among the various cases. However, our focus is on frozen precipitation retrieval, thus, we will emphasize the use of high frequency channels (90-200 GHz) since these are most sensitive to the ice in clouds. For comparison, improvement, and validation, we will implement our approach in our own retrieval scheme as well as in a combined/radiometer-only package made available by our Co-I. To help test and validate our algorithm, we will also use field data provided by collaborators and/or publicly available observational datasets.
The significance of our proposed work is that it will provide falling snow detection and estimation capabilities for the GPM core satellite and for use on other radiometers with millimeter-wave channel sets. Furthermore, our studies and approaches to address land surface emissivity will be applicable to other precipitation regimes such as light rain over land, which has been traditionally difficult to measure from space. Our estimates of snowfall will allow for measurements of frozen precipitation on a more global basis and a more complete understanding of the link between falling snow and hydrological processes. This proposal directly addresses the Precipitation Science (PMM) ROSES NRA item: "Development, testing, and validation of retrieval algorithms for high frequency (90-200 GHz) radiometer channels for improved estimation of light rain and snow over land and ocean" under the Retrieval Algorithms, Validation, and Multisatellite Precipitation Analysis topic area. The PI, Co-I's, and collaborators assembled for this study have demonstrated capabilities for snowfall retrievals, TRMM rain rate retrievals, measuring the physical properties of snow, and modeling the radiative properties of nonspherical ice and mixed-phase precipitation particles.
