Impact of Hybrid Vertical Coordinate to the WRF RAP-Based Upper-Level Turbulence Forecast

Unrealistically large areas of light-or-stronger turbulence are predicted via forecast vertical velocity over mountainous regions by the operational turbulence forecast system for the NOAA Aviation Weather Center product. This is derived from the post-processing of output from the operational WRF-ARW (Advanced Research Weather Research and Forecast) Rapid Refresh (RAP) model. It turns out that the over-forecasting is partly a consequence of using the terrain-following sigma coordinate in the WRF-RAP model. Especially over regions of steep mountain ranges, the terrain-following sigma levels depart significantly from the horizontal, even in upper troposphere and lower stratosphere (UTLS). This results in spurious horizontal and vertical gradients of wind fields in the model. When used as input to the operational turbulence forecast system, this gives high false alarm rates for light-or-stronger turbulence forecasts in the UTLS near the cruising altitudes of commercial passenger aircraft.

Partly motivated by this over-forecasting, a hybrid vertical coordinate based on the formulation of Klemp (2011, Mon. Wea. Rev.) has been implemented into the WRF-ARW by NCAR/MMM, and evaluated in both the RAP and the High-Resolution Rapid Refresh (HRRR, which also uses the WRF-ARW model) by NOAA/GSD. The NOAA/GSD started real-time parallel running of the updated WRF-RAP during March 2017.

The aim of this study is to directly compare the upper-level turbulence forecast product based on output from RAP with the new hybrid coordinate system with this product based on output from the current operational version of RAP using the traditional sigma coordinate, both validated against observed in situ turbulence reports over the Western mountainous US. As an example, a 6-h forecast initialized at 1200 UTC on 25 May 2017, using the current operational, version of the RAP predicts a noisy vertical motion field at 35,000 feet (~10.5 km above ground level) in vicinity of an upper-level jet that extends from over the Alaska Panhandle southeastward to the central Rockies. This vertical motion field is dramatically smoother in the RAP 6-h prediction initialized at the same time and using the hybrid vertical coordinate. Since the operational turbulence forecast system uses the model-predicted vertical motion fields as one of its inputs, using output from the new hybrid coordinate version of RAP gives a turbulence forecast consistent well with the null (smooth) intensity of turbulence reports from commercial aircraft over this region at this valid time. More detailed spectrum analyses and statistical evaluation against in situ Eddy Dissipation Rate (EDR) turbulence measurements from commercial aircraft will be presented in the conference.

From this we expect to see how the updated WRF-RAP with hybrid vertical coordinate (scheduled to become operational early in 2018) can give better performance of upper-level vertical velocity and, consequently, turbulence forecast over the mountainous regions of the US. This illustrates how research-to-operation (R2O) collaboration between development and operational teams can give a huge benefit for end users.