| description abstract | The spatiotemporal distribution of the vertical velocity at synoptic and subsynoptic scales is key to the patterns of weather and climate on earth. On these scales, the vertical velocity is on the order of one to a few centimeters per second, typically about three orders of magnitude smaller than typical horizontal wind velocities. Because of the smallness of large-scale vertical velocities relative to typical horizontal velocities, a direct observation of the large-scale vertical air velocity is extremely difficult. In a case study on observational material obtained during a 68-h experiment using the SOUSY very high frequency (VHF) radar in the Harz Mountains in Germany, the authors present the first intercomparison between three different sources of physical information that can provide large-scale vertical wind velocities: (i) the Doppler shifts observed with a vertically pointing VHF radar; (ii) the rates of change of the altitudes of refractive-index discontinuities as identified with frequency-domain interferometry (FDI), which is still a relatively unexplored technique in meteorology; and (iii) the output of a regional numerical weather prediction model (NWPM), which has been set up to model the meteorological situation during the observational period. There are several phenomena that have been known to possibly cause significant biases in mean vertical velocities retrieved from the Doppler shifts measured with vertically pointing clear-air VHF radars: (i) stationary or nonstationary gravity waves with vertical-velocity amplitudes up to the order of 1 m s?1; (ii) stationary or horizontally advected tilted refractive-index discontinuities that are aspect sensitive in the VHF regime; and (iii) a correlation between the radar-reflectivity fluctuations and the vertical-velocity fluctuations within a vertically propagating gravity wave. On the basis of an intercomparison between the vertical velocities retrieved from (i) ?standard Doppler? VHF radar observations, (ii) VHF FDI observations, and (iii) the NWPM output, the authors present first evidence that, under ideal conditions, VHF FDI can be used to directly monitor large-scale vertical motion. | |
