High resolution wind speed measurements with quadcopter UAS: calibration and verification in a wind tunnel with active grid

Abstract. As a contribution to closing observational gaps in the atmospheric boundary layer (ABL), the SWUF-3D fleet of unmanned aerial systems (UAS) is utilized for in situ measurements of turbulence. To date, the algorithm for wind measurement has only been calibrated in the free field. We therefore present in this work the calibration and verification under laboratory conditions. The UAS measurements are performed in a wind tunnel with active grid and constant temperature anemometers (CTAs) as a reference. Calibration is performed in x- and y-coordinate directions of the UAS body frame in wind speeds of 2…18 m s^-1. For systematic verification of the measurement capabilities and identification of limitations, different measurement scenarios like gusts, velocity steps and turbulence are generated with the active grid. Furthermore, the measurement accuracy under different angles of sideslip (AoS) and wind speeds is investigated and it is examined whether the calibration coefficients can be ported to other UAS of the fleet. Our analyses show that the uncertainty depends on the wind speed magnitude and increases with higher wind speeds, resulting in an overall root-mean-square error (RMSE) of less than 0.2 m s^-1. Applying the calibration coefficients from one UAS to others within the fleet results in comparable accuracies. Flights in different gusts yield an RMSE of up to 0.6 m s^-1. The maximal RMSE occurs in the most extreme velocity steps (i.e. a lower speed of 5 m s^-1 and an amplitude of 10 m s^-1) and exceeds 1.3 m s^-1. For variances below approx. 0.5 m² s^-2 and 0.3 m² s^-2, the maximal resolvable frequencies of the turbulence are about 2 Hz and 1 Hz respectively. We report successful calibration but with susceptibility to high AoS in high wind speeds, no necessity of wind tunnel calibration for individual UAS, and the need for further research regarding turbulence analysis.