Development of a Snowfall Retrieval Algorithm Using Data at High Microwave Frequencies

A snowfall retrieval algorithm is developed using high frequency satellite microwave data based on Bayes theorem. In this algorithm, observational data from the Wakasa Bay 2003 field experiment and the CREST's winter MCSs project in 2001 are used to construct a priori database of snowfall profiles. These profiles are then used as input to forward radiative transfer calculations to obtain brightness temperatures at high microwave frequencies. In the radiative transfer simulations, several size distributions for snowflake and many observed atmospheric sounding profiles are used. In addition, the scattering properties of the snowflakes are calculated based on realistic nonspherical shapes using discrete dipole approximation. The snowfall retrieval algorithm is validated for three snowfall cases in 14, 16 and 27 January 2001 over the west part of Japan. The retrieved snowfall rates from the Advanced Microwave Sounding Unit-B (AMSU-B) data are compared with surface radar observations. The results show a reasonably good agreement.