Considerations for WxUAS Wind Speed Estimation

Abstract Weather-sensing uncrewed aerial systems (WxUAS) have continued to advance to the point where they have become potential solutions for filling the critical data gap in the atmospheric boundary layer. WxUAS are unique compared to other UAS on the market since they need to operate in a large envelope of environmental conditions, and the conditions where they are most needed are often extreme. The WxUAS community has made significant progress in advancing thermodynamic sensing from WxUAS, but work is still needed to improve wind estimates over the full envelope of winds that WxUAS may face. This study evaluates three different methods for wind speed estimation, all which use the tilt of the WxUAS as it counteracts the wind. Each method is evaluated for accuracy and sensitivity to calibration using data from three separate field campaigns where the University of Oklahoma’s CopterSonde WxUAS and a Doppler wind lidar were co-located. The quadratic method proposed by Wetz and Wildmann (2022) was the most accurate overall, but was sensitive to calibration. The linear method (Palomaki et al. 2017), which has been historically used for the CopterSonde, was the least accurate but also least sensitive to the calibration dataset. On average, all three methods evaluated met the World Meteorological Organization ”Breakthrough” target of 3 m s −1 , and two met the ”Goal” target of 1 m s −1 .