ADVANCED TECHNOLOGY FOR SATELLITE MICROWAVE WATER VAPOR RETRIEVAL

OPERATIONAL APPROACHES FOR THE RETRIEVAL OF TEMPERATURE AND MOISTURE PROFILES FROM DMSP MICROWAVE AND MILLIMETER WAVE SOUNDER DATA ARE BASED ON EXPLOITATION OF THE SPECTRAL INFORMATION CONTENT OF THE DATA ALONE. REGRESSION RETRIEVALS ARE CONSTRAINED TO A SOLUTION SPACE DEFINED BY THE CLIMATOLOGICAL VARIANCE (CV) OF THE ENSEMBLE FOR WHICH THE STATISTICS ARE DEVELOPED. GLOBAL STATISTICW MAXIMIZE THE SAME SPACE AND VARIANCE. THIS INCREADES THE FAMILY OF POSSIBLE SOLUTIONS. TEMPERATURE AND MOISTURE PROFILES ARE OFTEN RETRIEVED FOR ASSIMILATION WITHIN METEOROLOGICAL APPLICATIONS MODELS WHICH THEMSELVES CAN PROVIDE INGORMATION ON THE ATMOSPHERIC STATE. AN EXAMPLE IS THEIR USE IN NUMERICAL WEATHER PREDICTION (NWP) MODELS. NWP MODELS DYNAMICALLY CONSTRAIN THE POSSIBLE STATES OF THE ATMOSPHERE (IE. FORECASTS OF TEMPERATURE AND MOISTURE PROFILES) THROUGH THEIR MODEL PHYSICS. THIS REDUCES THE SOLUTION SPACE FOR THE RETRIEVALS FROM THE CV OF THE REGRESSION SOLUTIONS. FURTHER CONSTRAINTS ARE POSSIBLE THROUGH PARAMETERIZATION OF THE PROFILE VARIANCES USING EMPIRICAL ORTHOGONAL FUNCTIONS (EOFS). THE OBJECTIVE OF THIS STUDY IS TO EXPLORE THE FEASIBILITY OF EMPLOYING PHYSICAL CONSTRAINTS FROM METEOROLOGICAL APPLICATIONS MODELS IN A SATELLITE DATA RETRIEVAL SYSTEM CONSISTING OF RETRIEVAL APPROACH AND APPLICATIONS MODEL PHYSICS. THE DESIRED OUTCOME IS TO IMPROVE THE ACCURACY OF PROFILES WITH SIGNIFICANT CLIMATOLOGICAL VARIABILITY SUCH AS WATER VAPOR.