Application of a high-resolution hourly-updated numerical weather prediction system for aviation-related cloud and icing forecasts: System refinement and associated model physics and data assimilation research

For the past two decades, NOAA’s Global Systems Laboratory (GSL) developed and refined a regional hourly-updated high resolution numerical prediction system known as the High-Resolution Rapid Refresh (HRRR).  This system is run operationally by NOAA and has been widely used for aviation, severe weather, renewable energy, and general short-range forecast applications.  Prediction of aviation-related hazards has always been a main focus area for this model system and much of the associated model physics and data assimilation research has targeted accurate prediction of aviation hazards such as convection, low clouds and fog, icing, and turbulence.  At present, GSL and the NOAA Environmental Modeling Center (EMC) are working together toward the implementation of a major upgrade to the HRRR, a new system known as the Rapid Refresh Forecast System (RRFS).  This new system retains many of the same model physics and data assimilation aspects as the HRRR, but uses a new model dynamic core, the FV3, developed at the NOAA Geophysical Fluid Dynamics Laboratory (GFDL), and also has a much larger 3-km domain than the HRRR as well as an ensemble forecast system. Key components of the RRFS model system for cloud and icing prediction include customized versions of the MYNN boundary layer scheme and the Thompson aerosol-aware microphysics (including a cloud faction variable), and a non-variational cloud analysis scheme to help with the initialization of clouds from METAR and satellite-based cloud observations.  Overall assessment of the RRFS vs. HRRR for cloud prediction skill, especially for aviation-related aspects (low ceilings, etc.) in the cold season, has revealed an improvement in the RRFS over the HRRR, which had a tendency to miss low stratus decks.  The RRFS has continued the good HRRR skill for icing prediction, as revealed by assessments associated with the 2019 ICICLE field project data sets.  As part of GSL’s aviation-related model work, grids from the HRRR have been provided to teams producing aviation-hazard specific guidance products for clouds and ceiling, icing, turbulence, and convection.  At the conference, we will describe the relevant model physics and data assimilation components of the RRFS system and present results for cloud and icing forecasts.