A Microwave-Based Hydrometeor Profile Retrieval Algorithm Using a Variational Technique S.-A. Boukabara, F. Iturbide-Sanchez, K. Garrett, W. Chen, C. Grassotti NOAA/NESDIS/STAR, 5200 Auth Rd, Camp Springs, MD 20746, Tel.: 301 763 8136 x 195 E-mail:sid.boukabara@noaa.gov 1. INTRODUCTION This study presents the results of a microwave-based algorithm which attempts to perform the retrieval of hydrometeor parameters in a profile form. The hydrometeors in question are both phases of the precipitating rain (liquid and frozen). The algorithm is called the Microwave Integrated Retrieval System (MiRS) which uses microwave radiances from AMSU and MHS sensors onboard NOAA-18, NOAA-19, Metop-A and DMSP-F16 SSMIS. It is running operationally at NOAA/NESDIS for a number of atmospheric, land and cryospheric products as well as for the integrated ice water path and surface rainfall rate. The retrieval is performed in a reduced space to ensure inversion stability, using Empirical Orthogonal functions (EOF). The assessment of the results is performed by using TRMM/TMI-PR profiles to ensure that vertical cross sections (global and regional) are consistent. Given the footprint sizes difference, the time synchronization issues, and the nature of the phenomena being sensed (rapid transformation/motion of the hydrometeors) it is shown that collocation is a major challenge. A discussion will summarize the final results that will be obtained in this on-going study. 2. ALGORITHM APPROACH MiRS is a One-Dimensional Variational inversion scheme (1DVAR) that employs the Community Radiative Transfer Model (CRTM) as the forward and adjoint operators. It solves simultaneously for the surface and the atmospheric parameters including hydrometeors. The iterative process of the 1DVAR inversion scheme employed in this study aims at minimizing the following cost function, similar to the variational radiance data assimilation employed in Numerical Weather Prediction (NWP) models: サ シ コ ォ ャ ェ u u サ シ コ ォ ャ ェ u u ) ( ) ( 2 1 2 1 ) ( 1 0 1 0 X Y Y E X Y Y X X B X X X J m T m T