Realtime WRF LES Simulations to Support UAS Flight Planning and Operations During 2018 LAPSE-RATE

Abstract. The simulations were forced using data from the High-Resolution Rapid Refresh (HRRR) and the Global Forecast System (GFS) obtained from the National Center for Environmental Prediction (NCEP) Central Operations (NCO). The nested WRF configuration used in this study featured a large-eddy-permitting, 111-m grid spacing mesh for the inner-most domain. Output from the WRF forecasts were processed on-the-fly using the Unified Post Processor (UPP). A THREDDS data server was coupled with a web-viewer to provide real-time graphics that were used to support UAS flight planning and decision making. The simulations ran in under six hours on the National Center for Atmospheric Research (NCAR) Cheyenne supercomputer using 59 cores (2124 processors). The simulations were run twice per day (12 runs total during the experiment) to support both next-day mission planning and day-of tactical guidance. The simulations provided a high-resolution depiction of the four-dimensional variability of weather conditions across the northern half of the San Luis Valley, Colorado, where UAS operations took place. The simulations were used to assess the possibility of a number of small UAS weather hazards including wind shift boundaries, vertical shear, strong thermals, turbulence intensity, fog, low ceilings, and thunderstorms. Details of the model configuration used to perform the simulations and the data processing steps used to produce the final grids of state variables and other sensible weather products (e.g., ceiling and visibility, turbulence) are given. A few examples of predictive capabilities of the modeling system are provided to illustrate the skill of the model at predicting fine-scale boundary layer structures and turbulence associated with drainage winds, up-valley flows, and convective storm outflows. A subset of the at data is available at the Zenodo data archive (https://zenodo.org/communities/lapse-rate/) while full dataset (see Pinto et al. 2020a) is archived on the NCAR Digital Asset Services Hub (DASH) and may be obtained at https://doi.org/10.5065/83r2-0579.