Developing a Unified Prognostic Cloud Scheme for the UFS

Many physics suites used in today’s NWP models still contain inconsistencies across the spectrum of parameterized cloud and precipitation processes, which may be contributed to by the microphysics, convection, and/or boundary layer schemes. For example, subgrid-scale clouds may not be able to precipitate, may only be represented in convective regimes, and typically have only a subset of the species represented in most microphysics schemes. In addition, some suites carry more than one cloud fraction or apply cloud fractions inconsistently in their coupling with radiative transfer. This study is an attempt to develop a physically coherent representation of all cloud and precipitation processes for future versions of NOAA’s Unified Forecast System. The starting point is GFDL’s prognostic (Tiedtke) cloud scheme as implemented in the AM4 climate model, which has been ported into the CCPP and has initially been integrated into an experimental version of the HRRR suite. Initial coupling work has been performed with a modified form of the Mellor–Yamada–Nakanishi–Niino (MYNN) Eddy Diffusivity-Mass Flux (EDMF) scheme, as well as the Thompson microphysics scheme. Modifications include augmenting the large-scale prognostic clouds produced by the Tiedtke scheme with MYNN EDMF-produced subgrid-scale clouds and accounting for subgrid-scale partial cloudiness in the microphysics. This presentation will overview the design, research goals, and preliminary results, focused both on single-column model results and 3D testing. Sensitivities, successes, and remaining challenges are identified. Despite this project not being complete, this presentation can still serve as a discussion-point for the future of fully-integrated moist-turbulent physics in NWP models.