Cloud and Precipitation Representation in the Marine Boundary Layer Using the Common Community Physics Packaged (CCPP) Single Column Model (SCM)

Assessments of retrospective forecasts using the pre-operational Global Forecast System (GFS) v16 provided by the NOAA Environmental Modeling Center, and physics innovations provided by developers, were conducted as part of the Unified Forecast System (UFS) Research to Operations (R2O) project, with the goal of improving forecast skill of NOAA’s operational models. Diagnosed physical properties from the year-long assessments include systematic errors and biases in cloud coverage, precipitation, and radiative fluxes which are associated with problematic cloud and microphysical parameters, i.e. cloud liquid and ice water paths and their partitioning. The results suggest that uncertainties in cloud-radiation interactions and their sensitivity to the model physics remain an ongoing issue in medium-range weather forecasts. In particular, an underestimation of marine stratocumulus clouds remains in the GFSv16 during the boreal summer off the west coasts of continents.

Using the CCPP SCM with the operational GFS version 16 and physics updates for the future versions of the operational GFS (i.e prototype-8 and high resolution-1) at varying grid spacings, a hierarchy of tests was conducted to gain insight into the key physical processes related to biases in macrophysical, microphysical, precipitation, and radiative properties. Various sensitivities include exploring contributions of the individual components of large-scale forcings used to drive the SCM, horizontal entrainment rate in shallow cumulus, and effects of cloud concentration nuclei (CCN) amount on precipitation. A summer case spanning 20-25 July 2013 (Leg 15A) from the Marine ARM GPCI Investigation of Clouds (MAGIC) campaign, exhibiting a classic stratocumulus-to-cumulus (Sc-to-Cu) transition over the eastern North Pacific is investigated, where the misrepresentation of marine stratocumulus clouds can largely affect net radiative forcing and are crucial in the dynamics of air-sea coupled systems. While global forecasts exhibited a negative bias in total cloud fraction, SCM simulations of the aforementioned case did not underestimate but rather produced slower growth, coupled with issues of boundary layer decoupling and Sc-to-Cu transition, compared to campaign observations. While slow growth rate can be associated with underpredicted cloud-top entrainment as a result of cold, moist biases and stratification strength, sensitivities of the entrainment rate value and CCN amount did not show improvements in the representation of clouds and precipitation. Sensitivities in large-scale forcings showed that when only potential temperature advection was considered, the Sc breakup was more comparable to campaign observations, suggesting other large-scale forcings may negatively contribute to the biases. Further evaluation of clouds, precipitation, and their interactions will be extended to include additional MAGIC cases with continued testing across pre-operational physics updates and innovations. Key findings will be presented.