The retrieval and validation of Doppler velocity estimates from range imaging

Abstract

Range imaging (RIM) is used to describe a constrained optimization signal-processing method that can be applied to wind profilers capable of operating over a small set of distinct transmitter frequencies. The results of the signal-processing method are typically high-resolution maps of the backscattered power as a function of range. In this paper it is discussed how RIM processing can be implemented in order to additionally obtain high-resolution estimates of the Doppler velocity. The method has been demonstrated using data from a 915-MHz tropospheric profiler located in Platteville, Colorado. Examples of data collected during an experiment conducted on 10 April 2001 are presented. In this experiment the radar was operated alternately in two different modes. The cycle time for the two modes was about 50 s. The particular operation of the radar allowed comparison of radar reflectivity (η) and vertical velocity (V) measurements collected using the two modes. In the first mode, 2-μs pulses were transmitted and RIM processing was used to produce estimates of η and V on a grid with a separation of only 15 m. Without RIM processing the range resolution of the data would have been 300 m. In the second mode, 0.5-μs pulses were used, corresponding to a range resolution of 75 m. Estimates of η and V were then obtained from these data through conventional Doppler spectral processing. A conditional averaging method was used to process the reflectivity and vertical velocity data from the two modes. It is shown that the RIM-processed data can be used to resolve structures in the height profiles of η and V on scales less than those of the conventional range resolution of the radar as dictated by the pulse width. © 2004 American Meteorological Society.