Spaceborne passive microwave measurement of snowfall over land

A physically based retrieval algorithm was developed to estimate snowfall over land. The retrieval algorithm relies on the MM5 model that generates the vertical structure of a snow cloud, including snow mass, snow particle effective diameter, and water vapor. The MM5 cloud simulation was used to provide statistics for generating the cloud characteristics. The snow cloud profile and surface emissivity were then used in radiative transfer calculations that were optimized against AMSU-B observations at 89, 150 and 183.3±7, ±3, and ±1 GHz. The multi-parameter cloud model that produced brightness temperatures that best fit the AMSU-B observations was selected as the retrieved profile. The retrieved snowfall distribution was validated with radar reflectivity measurements obtained from the operational NWS radar network.