Using Multitaper to Improve Meteorological Estimates When Filtering Ground Clutter

One use of spectral analysis in weather radar is the removal of narrowband signals such as ground clutter to enable the recovery of underlying meteorological signals. In the spectral domain, these ground clutter signals may leak into adjacent frequencies (referred to as spectral leakage) and, if strong enough, can mask the meteorological signals. To contain the spectral leakage from ground clutter, high dynamic range windows are used, but their use also increases the variance of meteorological estimates. Instead of using conventional windows, we examine the multitaper technique for spectral-leakage mitigation. This approach lowers the variance of meteorological estimates compared to typical windows but does increase the computational complexity. To test the effectiveness of the multitaper technique, we integrate it into the CLEAN-AP filter, a spectral ground clutter filter that currently uses standard windows. Simulations combining both ground clutter and meteorological signals are used to quantify the reduction in the variance of recovered meteorological estimates and the improvement in the overall performance of CLEAN-AP.