An Improved M-PRT Technique for Spectral Analysis of Weather Radar Observations

摘要

The exploitation of Doppler radars for weather observations is strongly constrained by the well-known range–velocity dilemma. To overcome the range and velocity ambiguities, dual and triple staggered pulse-repetition time (PRT) techniques are commonly used in Doppler radar systems. Today, a triple-PRT (3-PRT) scheme is operational in France. These techniques imply nonuniform sampling of the weather signal, inducing multiple replicas in the Doppler spectrum. The situation is particularly complicated for short-wavelength radars, where larger extension factors of the unambiguous Nyquist interval are needed. To overcome these difficulties, a novel technique called OptM-PRT is proposed. It mainly consists in optimizing the transmission scheme based on multiple pulse repetition time, so that the corresponding autocorrelation function is well filled. The Doppler spectrum is therefore reconstructed with much less ambiguities, from the computation of the autocorrelation function of radar signal and its Fourier transform. Considering both 3-PRT and Opt9-PRT schemes, the magnitude and Doppler velocity of radar returns in rain are simulated for different spectral widths, with and without elimination of the spectral lines of ground clutter. When the ground clutter is filtered out, the 3-PRT is found to better reproduce the Doppler velocity, whereas the Opt9-PRT better restitutes the magnitude of the signal. In the presence of noise, the Opt9-PRT scheme produces the best result for both the magnitude and velocity. The 3- and Opt9-PRT techniques have been applied to the C-band Doppler radar operating in Bourges, France. The experimental results show that Opt9-PRT efficiently reconstructs the Doppler spectrum of rain echoes.