Impacts of Vertical Enhancements to NCAR's Prototype Current and Forecast Icing Products

The Current and Forecast Icing Product (CIP and FIP) provide hourly diagnoses and forecasts of in-flight icing conditions. These products are developed at the National Center for Atmospheric Research (NCAR) and transferred to the National Oceanic and Atmospheric Administration’s National Weather Service for operational implementation. The current operational CIP and FIP icing fields are derived from The Weather Research and Forecasting - Rapid Refresh (WRF-RAP) Numerical Weather Prediction (NWP) model forecast data with a horizontal grid spacing of 13 km. In addition to the forecast fields, CIP also incorporates lightning, satellite, METAR, low-level composite radar reflectivity, and voice pilot report observations to create its icing diagnosis. In the coming years, the operational CIP and FIP will transition to using the Rapid Refresh Forecast System (RRFS) version 1 NWP model, which has a horizontal grid spacing of 3 km. In preparation for this change, NCAR continues to develop a prototype CIP and FIP (P-CIP/FIP) to leverage experimental forecasts from the RRFS. Beyond changing the NWP model and the horizontal grid spacing, the P-CIP/FIP have undergone several advancements from the current operational CIP/FIP. These include switching to three dimensional radar reflectivity data and adding geostationary satellite channels in the P-CIP, increasing the dependence on NWP model forecasts of supercooled liquid water (SLW) in P-CIP/FIP, and a redefining of the meteorological scenarios used in the P-CIP/FIP to compute the Supercooled Large Drop (SLD) Potential output field.

One of the most significant recent advancements for the P-CIP/FIP has been to modify the algorithms to work on the native-hybrid vertical levels available from the RRFS. Historically the CIP and FIP have interpolated NWP forecast fields from these levels to isobaric levels with a 25 hPa vertical spacing. On several occasions, interpolation of the liquid water content fields has resulted in erroneous forecasts of SLW. This has been a particular issue near the melting level in areas of deep clouds near large winter cyclones. Additionally, the RRFS model has increased the number of vertical levels by 30% over the WRF-RAP (and also High Resolution Rapid Refresh, HRRR) from 50 to 65. As P-CIP/FIP research evolves, particularly in the area of SLW drop size, it becomes increasingly important to retain as much NWP model information as possible. Thus, the vertical interpolation to isobaric levels will be eliminated. In various places throughout the P-CIP/FIP algorithms there are assumptions built in that the NWP forecast fields have this vertical spacing. These assumptions were addressed by modifying the algorithms to use a physical layer thickness rather than an assumed isobaric thickness. Additional modifications will be discussed, and results from P-CIP/FIP using NWP model fields on isobaric levels will be compared to results using the native-hybrid level NWP model fields using both icing pilot reports, and airborne in situ data from a flight during the WINTRE-MIX field campaign.

This research is in response to requirements and funding by the Federal Aviation Administration (FAA). The views expressed are those of the authors and do not necessarily represent the official policy or position of the FAA.