Evaluation of and Suggested Improvements to the Morrison Microphysics in WRF Model -Part I: Analysis on Triple-Moment Size Distribution Function and Sedimentation Section

Observed rain droplet size distribution (DSD) could be employed to evaluate fitting ability of assumed size distribution functions (SDF) in bulk microphysics schemes, such as M-P Distribution and Gamma Distribution. Former observation studies used look-up table or fixed moment group methods to solve parameters in M-P SDF and Gamma SDF in order to fit observed DSDs. But these methods could hardly be applied into microphysics scheme directly, because only source/sink terms of zeroth and third order moments are possibly simulated in microphysics schemes. So, this study evaluated multiple parameter-solving methods available to microphysics schemes and quantified reliabilities of each ones. By comparison, more than 50% relative error for first to sixth order moments caused by M-P SDF could be modified, when it was replaced by a Gamma SDF with a diagnosed shape parameter. The improvement could be up to over 75% when solved shape parameter methods were used. These advantages of Gamma SDF showed requirement and possibility to add the third predicted parameter, namely the shape parameter, into SDF. Further, sedimentation section in Morrison Scheme was divided from its source code and converted into a triple-moment version. Then both original and converted versions were evaluated based on DSDs observed on three different heights. By analyzing source/sink terms and some modification, modeled number mixing ratio could be in same magnitude range with observed data. Summarized experiment results in this paper could serve as suggestions about how to develop and improve a triple-moment Morrison Microphysics Scheme.