Downscaling wind data from a mesoscale numerical weather prediction model (HRRR) to a microscale fast-response wind modeling platform (QES-Winds)

Authors: BOZORGMEHR, BEHNAM - Washington State University; MARGAIRAZ, FABIEN - University Of Utah; STOLL, ROB - University Of Utah; Mahaffee, Walter; Lee, Jungmin; LIU, HEPING - Washington State University

Submitted to: American Meteorological Society
Publication Type: Abstract Only
Publication Acceptance Date: 3/10/2025
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Addressing challenges such as airborne pathogen transport, wildfire smoke exposure on grapes, and pollution dispersion in agronomic and urban environments require accurate high-resolution faster than real-time wind field modeling. Quick Environmental Simulation (QES) is a microclimate simulation platform designed to compute 3D environmental scalars in urban areas and over complex terrain. A key component, QES-Winds, is a fast-response 3D diagnostic wind model that prioritizes computational efficiency by solving a mass-conservation equation instead of employing slower, momentum-conserving solvers. The model generates initial wind fields using data from observations or mesoscale numerical weather models like HRRR (High-Resolution Rapid Refresh), leveraging interpolation techniques and empirical parameterizations. QES-Winds downscales HRRR’s 3 km grid resolution to a finer 100 m resolution, effectively capturing topographic and local wind effects subgrid to HRRR simulations. To enhance QES-Winds' predictive accuracy, improvements were introduced to estimate atmospheric stability through Monin-Obukhov length calculations. Stability conditions are derived from the Pasquill-Gifford classes, incorporating near surface wind speed, solar radiation and cloud cover data to create unstable, neutral, and stable wind profiles. In addition, the Monin-Obukhov length is directly estimated using surface momentum and buoyancy fluxes from HRRR to have more ways to include stability condition in the platform. Computational efficiency is further improved by adopting bilinear interpolation, which reduces runtime while preserving local data structures, making the model more scalable for larger domains. Validation against field data collected in the Yakima Valley, Washington, during the 2020 wildfire season demonstrated significant improvements in wind field accuracy compared to previous versions. QES-Winds simulations are better aligned with measured data, outperforming HRRR’s 3 km resolution in capturing local wind dynamics. These advancements establish QES-Winds as a practical and efficient tool for high-resolution wind field simulations, with future plans to integrate HRRR smoke data into QES for real-time smoke exposure analysis.