THE OVERALL OBJECTIVE OF THIS PROPOSAL IS TO EVALUATE QUANTIFY UNDERSTAND AND POTENTIALLY MITIGATE ERRORS AND UNCERTAINTIES THAT ARE PRESENT IN GPM S PASSIVE MICROWAVE RAINFALL RETRIEVAL ALGORITHM (GPROF) OVER MOUNTAINOUS TERRAIN. WHILE IT IS WELL UNDERSTOOD THAT OROGRAPHIC PRECIPITATION REGIMES POSE PARTICULAR CHALLENGES FOR BOTH ACTIVE AND PASSIVE MICROWAVE INSTRUMENTS THERE ARE THREE FUNDAMENTAL QUESTIONS THAT AFFECT GPROF S PERFORMANCE IN OROGRAPHY THAT ARE NOT SUFFICIENTLY WELL UNDERSTOOD TO IMPLEMENT IMPROVEMENTS AT THIS TIME. THEY CAN BE SUMMARIZED AS: (1) DOES THE A-PRIORI DATABASE CONSTRUCTED FROM THE GPM RADARS PROPERLY CAPTURES THE SHALLOW PRECIPITATION OFTEN ASSOCIATED WITH OROGRAPHY? (2) ARE THE SCATTERING SIGNALS THAT THE RADIOMETER ALGORITHM RELY ON TO INFER PRECIPITATION CONSISTENT WITHIN DISTINCT DYNAMICAL REGIMES? (3) CAN THESE DYNAMICAL REGIMES AND ASSOCIATED MICROPHYSICS BE INFERRED FROM ANCILLARY DATA? WE PROPOSE HERE TO EMBARK UPON A SYSTEMATIC EXAMINATION OF OROGRAPHIC PRECIPITATION OVER THE WESTERN UNITED STATES WHERE ADDITIONAL DATA IS AVAILABLE TO GUIDE THE RESEARCH. WE WILL BEGIN BY IDENTIFYING GROUND BASED RADARS ACROSS THE US MOUNTAIN WEST WHERE THE QUALITY INDICATORS ARE HIGH FOR BOTH RAIN AND SNOW AND WHERE UNCERTAINTIES CAN BE ASSESSED. THE GROUND BASED RADAR ESTIMATES THEMSELVES WILL BE COMPARED TO BOTH PRISM AND SNOWTEL DATA ON DAILY AND LONGER TIME PERIODS TO BOTH CHARACTERIZE ACCURACY AND QUANTIFY UNCERTAINTIES. THE OLYMPEX FIELD EXPERIMENT DATA WILL BE INCLUDED IN THIS SET AS IT PROVIDES AN EVEN RICHER DATA SET TO COMPARE AGAINST. ONCE WE HAVE ESTABLISHED BASELINE LOCATIONS FOR VALIDATING INSTANTANEOUS SATELLITE PRECIPITATION ESTIMATES WE WILL EXPLORE QUESTION (1) ABOVE NAMELY THE EXTENT TO WHICH THE DPR AND COMBINED ALGORITHMS CAPTURE THE PROPER RAINFALL AND SNOWFALL STRUCTURES IN DIVERSE METEOROLOGICAL REGIMES. AS THE USE OF REGIMES IS OFTEN SUBJECTIVE FOR OBJECTIVE (2) WE WILL USE THE HIGH RESOLUTION RAPID REFRESH (HRRR) MODEL THAT IS RUN OPERATIONALLY OVER THE CONTINENTAL UNITED STATES WITH 3KM RESOLUTION AND A 1 HOUR REFRESH RATE. THE HRRR MODEL WILL BE QUERIED SPECIFICALLY FOR REGIMES OF SIMILAR CLOUD VERTICAL DEVELOPMENT IN THE VICINITY OF THE OBSERVED PRECIPITATION. WITHIN THESE REGIMES WE WILL USE THE HYDROMETEORS RETRIEVED FROM THE GPM RADARS AS WELL AS INFERRED FROM GROUND-BASED RADARS TOGETHER WITH RADIATIVE TRANSFER SIMULATIONS TO ASSESS IF THEY PROPERLY CAPTURE THE SCATTERING SIGNAL SEEN IN GMI. FINALLY OBJECTIVE (3) WILL BE ADDRESSED BY COMPARING THE GLOBAL ANALYSES TO SEE IF THE REGIMES IDENTIFIED LOCALLY BY HRRR CAN BE EXPANDED TO INCLUDE THE ENTIRE GLOBE WHERE NO CLOUD RESOLVING MODEL SIMULATIONS ARE AVAILABLE. THE OLYMPEX DATASET WILL BE USED ALONG THE WAY TO ENSURE THAT EACH STEP AND ASSUMPTION CAN BE PHYSICALLY CONNECTED WITH THE ACTUAL CLOUD STRUCTURES AS OBSERVED IN THE FIELD PROGRAM. WHILE THE ABOVE WORK IS FOCUSED ON IMPROVING A PRODUCT THAT IS KNOWN TO BE DEFICIENT IN MOUNTAINOUS TERRAIN WE WILL SIMULTANEOUSLY USE THE REGIME CLASSIFICATION ADDRESSED IN QUESTION (2) TO DEVELOP A MORE SOPHISTICATED UNCERTAINTY ESTIMATE THAT REPORTS NOT ONLY THE INDEPENDENT PIXEL UNCERTAINTIES BUT ALSO THOSE UNCERTAINTIES COMING FROM UNKNOWN OR UNCERTAIN REGIME CLASSIFICATIONS THAT CAN BE SIGNIFICANT AND HIGHLY CORRELATED BETWEEN PIXELS OVER A GIVEN AREA.