9.2      Improving cloud and solar radiation forecasts in the RAP/HRRR forecast systems.

Olson, Joseph B., Jaymes S. Kenyon, NOAA/CIRES, Greg Thompson, NCAR, John M. Brown, NOAA, Wayne M. Angevine, NOAA/CIRES, Dave Turner, Stan Benjamin, and Georg Grell, NOAA

The 13-km Rapid Refresh (RAP) and 3-km convection-allowing High-Resolution Rapid Refresh (HRRR) are hourly-updating operational forecast models that support short-range forecasting interests for the United States. Among the many important forecast variables/quantities, cloud ceiling and downward shortwave radiation are of primary importance for the aviation and energy industry, respectively. Both quantities require skillful prediction of resolved-scale clouds from the Thompson microphysics scheme, subgrid-scale clouds from the Mellor–Yamada–Nakanishi–Niino (MYNN) Eddy Diffusivity-Mass Flux (EDMF) moist-turbulence parameterization scheme, and proper coupling to the RRTMG longwave and shortwave radiation schemes.

Model physics development within the RAP/HRRR has focused on improving the representation of shallow cumuli and stratiform subgrid clouds to compliment the resolved scale clouds. The macro- and microphysics within the MYNN-EDMF scheme have been improved for WRF-ARWv4.1, with more accurate subgrid mixing ratios, cloud fractions, and effective radii. The combined impact of these improvements are shown to drastically reduce biases in downward shortwave radiation at the surface, improve the skill of cloud ceiling forecasts, as well as improve the general forecast skill of the RAP and HRRR.

This talk overviews these developments and presents some results of these achievements both in the context of case studies and longer-term retrospective hourly cycled tests. We will also discuss some of the remaining challenges specific to cloud forecasting as well as the consequences of needing to re-tune other aspects of the forecast system after improving these long-standing historical biases.