Estimating meteorological wind vector components from monostatic Doppler radar measurements: A case study

Abstract

During the last decade, great progress has been made in the use of sensitive VHF and UHF Doppler radars (often called MST, ST, or wind profiler radars) to study the dynamics of the clear air in the troposphere and middle atmosphere. Currently used two‐beam and three‐beam metods of transforming observations of radial velocity υr from these radars into wind vector components u, υ, and w were tested for consistency in a five‐beam, 4‐hour case study. The study was made at the Sunset radar field site located 14 km west of Boulder, Colorado, and 17 km east of the continental divide, in the foothills of the Rocky Mountains. A weak jet stream was overhead. Under these conditions, the commonly used two‐beam relation (which assumes horizontal uniform flow) for transforming υr to u and υ gave grossly biased results (systematic discrepancy larger than 5 m/s). The three‐beam relation, based on three‐dimensional uniform flow, greatly reduced the error for much of the observation period. However, localized large errors remained at some heights and times, implying the need for a technique based on a three‐dimensional linearly varying flow field. Examination of a 16‐day comparison of vertical velocity observations made at the Sunset (mountains) and the Platteville (plains, 75 km east of the continental divide), Colorado, radar sites indicated that the flow away from orographic influences is simpler but that the two‐beam transformation may still lead to biased results much of the time.