Search Results

You are looking at 1 - 2 of 2 items for :

  • Author or Editor: Michael Tjernström x
  • Journal of Atmospheric and Oceanic Technology x
  • All content x
Clear All Modify Search
Michael Tjernström and Patrick Samuelsson

Abstract

Wind calculations from a radome gust probe system, in combination with a wander-angle Inertial Navigation System (INS), are used to calculate turbulence (variance and fluxes) by eddy correlation. The difference in time response between the air motion gust probe and the INS is analysed in postprocessing of in-flight data from calibration maneuvers. The effect of this difference on mean winds during accelerated maneuvers and on turbulence spectra and cospectra from low-level turbulence measurement flight legs is demonstrated.

The estimated time delay between the two systems is small for most parameters, less than 0.1 s, but is large for ground speed, about 0.1 s, and substantial for true heading, almost 0.5 s. In forced maneuvers (pitching, yawing, and turning), the effect on the calculated mean wind is significant. During typical turbulent measurement conditions, that is, straight and level flight, the effect on the winds and thus on spectra and cospectra is, however, small.

Full access
Michael Tjernström and Carl A. Friehe

Abstract

A “radome gust probe” system was installed on a twin-jet aircraft for the purpose of boundary-layer research. This system provided a useful relatively low-cost method for air motion and turbulence measurements on an aircraft already equipped with an inertial navigation system (INS) and a data acquisition system. An error analysis was made for the wind measurements and gave the limitations for the present system with an unmodified airliner-type INS, The major factors that limit the precision of the horizontal wind are the resolutions and accuracy of the aircraft ground speed components and the true heading. A simple method was devised to improve the heading resolution. From in-flight maneuvers, it was determined that the mean horizontal airspeed vector was accurate to <0.5 m s−1—limited by the long-term drift and oscillation errors from the INS-and that pitch and yaw contamination of the wind was less than 5%. The in-flight data indicate that there probably are unknown time lags within the INS, which degrade the wind vector measurement for anything but straight and level flight. Some intercomparisons were obtained from fly-by(s) past an instrumented 140-m tower. These showed general agreement between mean, variance and high-frequency spectral measurements of the velocity components and temperature. Flux or covariance comparisons were not as good, probably due both to the short flight tracks and a complex boundary layer structure at the tower.

Full access