WRF wind field assessment under multiple forcings using spatialized aircraft data. (28th September 2020)

• Record Type: Journal Article
• Authors: Carotenuto, Federico; Gualtieri, Giovanni; Toscano, Piero; Miglietta, Franco; Gioli, Beniamino
• Is Part Of: Meteorological applications. Volume 27:Number 5(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-09-28
• Subjects: aircraft observations -- CFSRECMWF -- modellingwind -- WRF; Meteorology -- Periodicals; Meteorological services -- Periodicals; 551.5
• DOI: 10.1002/met.1920
• Languages: English
• ISSNs: 1350-4827

Abstract

Abstract: The performances of limited area weather models are affected by the choice of core solvers, domain resolutions, and initial and boundary conditions. To understand the extent of such differences on simulated wind fields, weather research and forecast (WRF) simulations initialized by different forcings were extensively compared with an aircraft‐derived high‐resolution data set. The two used forcings were the European Centre for Medium‐Range Weather Forecasts (ECMWF) ERA‐Interim reanalysis and the National Centers for Environmental Predictions (NCEP) Climate Forecast System Reanalysis (CFSR). The model domain covered a large portion of central western Italy (including part of the Tyrrhenian coast) encompassing the aircraft track and allowed the characterization of their performance across the simulation domain rather than a small set of point‐based observations. The WRF results show good agreement with the aircraft data across the whole flight track with both forcings (root mean square errors (RMSEs) < 2.3 m·s −1 and an average r 2 = 0.7). Orography and coasts show an effect on simulated wind fields. The presence of a strong orography (which is smoothed by the model internal terrain elevation model) is associated with increased errors. Distance from the coast is also associated with a variation in RMSE (even if in a non‐straightforward manner) because of potential breeze effects. No forcing data set clearly outperforms the other, while the ECMWF has higher correlation co‐efficients when considering wind direction. Abstract : A research aircraft, such as the one shown, is capable of producing precious data sets of wind speed and wind direction at high frequency and over different land uses and orography. Such a data set allows a more thorough evaluation of the numerical weather models' simulations compared with sparse fixed ground stations. The evaluation shows that the model is quite capable of representing such highly spatialized data, and that the choice of initial forcing has only a moderate impact.