A Comparison of FPS-16 and GMD-1 Measurements and Methods for Processing Wind Data

Abstract

A machine method for computing winds from radiosonde and GMD-1 rawin tracking data is developed and the resulting winds are compared to those computed from precision FPS-16 radar tracking data. A direct comparison is possible because the measurements were made simultaneously by tracking the same effective target, which for the FPS-16 was metallic chaff in the balloon that carried the radiosonde transmitter. With GMD measurements 10 times min?1 instead of the usual 1 min?1 and when elevation angles are >10°, the computed winds and ascent rates reproduced both the macroscale and mesoscale features derived from the FPS-16 data. The mesoscale features which have vertical wavelengths from 1.5?3 km include oscillations in wind direction and speed. When the elevation angles are <10°, the GMD antenna is misdirected by indirect signals reflected from the ground and large spurious oscillations and steps appear in the elevation angle measurements. It is impossible to recover the mesoscale winds from these erratic measurements, but the macroscale or synoptic scale winds can be obtained by fitting a low order polynomial to the entire set of low elevation angle measurements. The winds computed from the FPS-16 measurements contain, between the earth's surface and the maximum altitude reached by the radiosonde balloon, many oscillations in wind speed and direction. The hodographs suggest that the velocities can be decomposed into a mean wind vector and a perturbation vector which rotates cyclonically or anticyclonically with height. In the stratosphere, anticyclonic rotation predominates. The perturbation vectors have magnitudes ranging from 0.5?10 m sec?1 and rotate through 2π rad in 0.5?3 km.