Using a Regression Ground Clutter Filter to Improve Weather Radar Signal Statistics: Theory and Simulations

Abstract

Ground clutter filtering is an important and necessary step for quality control of ground-based weather radars. In this paper, ground clutter mitigation is addressed using a time-domain regression filter. Clutter filtering is now widely accomplished with spectral processing where the times series of data corresponding to a radar resolution volume are transformed with a discrete Fourier transform after which the zero and near-zero velocity clutter components are eliminated by setting them to zero. Subsequently for reflectivity, velocity, and spectrum width estimates, interpolation techniques are used to recover some of the power loss due to the clutter filter, which has been shown to reduce bias. The spectral technique requires that the in-phase (I) and quadrature (Q) time series be windowed to reduce clutter power leakage away from zero and near-zero velocities. Unfortunately, window functions such as the Hamming, Hann, and Blackman attenuate the time series signal by 4.01, 4.19, and 5.23 dB for 64-point times series, respectively, and thereby effectively reduce the number of independent samples available for estimating the radar parameters of any underlying weather echo. In this paper, a regression filtering technique is investigated, through simulated data, that does not require the use of such window functions and thus provides for better weather signal statistics. In a follow-on paper that is in preparation the technique will be demonstrated using both S-band polarimetric radar (S-Pol) and NEXRAD data. Here, it is shown that the regression filter rejects clutter as effectively as the spectral technique but has the distinct advantage that estimates of the radar variables are greatly improved. The technique is straightforward and can be executed in real time.