Calibration of a Radome-Differential GPS System on a Twin Otter Research Aircraft for Turbulence Measurements

Abstract

A five-hole radome pressure probe at the nose of a small two-engine newly instrumented research aircraft was combined with global positioning system (GPS) receivers in differential mode to obtain high frequency measurements of the wind vector in the atmospheric boundary layer with possible accuracy (root-mean-square error) of about 0.1 m s?1. This low cost and simple system can provide wind velocity measurements of sufficient accuracy to estimate turbulent fluctuations. Special aircraft maneuvers above the atmospheric boundary layer were used to calibrate the radome probe. The analysis of these data showed that the static pressure defect has a significant dependence on flow angles and is affected by the propellers when significant thrust is applied. Using a simple method, the authors found that the pressure distribution on the radome deviated from the one expected for airflow incident on a sphere by more than 5%, the authors also detected a problem in the attack angle differential pressure sensor. The calibration of the local attack and sideslip flow angles due to flow distortion by the aircraft was obtained using two different methods. The first method was a least wind variance one assuming a linear form for the calibration of flow angles. This method is easy to use and can be applied in the presence of turbulence, but does not reveal any possible nonlinear dependence or problems in the data. The second method was a direct one that assumes near?zero mean vertical wind velocity above the boundary layer, while an average horizontal wind was estimated using the airstream speed with respect to the aircraft and the aircraft velocity from the differential GPS data. These methods gave similar results and, thus, increased the reliability of the calibration. The performance of the calibration procedure of the whole system was tested by examining the sensitivity of estimated wind components to the aircraft motion (about 5%) and the quality of mean profiles and turbulence statistics in the boundary layer.