Abstract
Spaced antenna baseline wind retrievals, in conjunction with traditional Doppler measurements, are a potential means of fine angular resolution weather radar wind vector retrieval. A spaced antenna implementation on an X-band active phased array architecture is investigated via Monte Carlo simulations of the backscattered electric fields at the antenna array. Several retrieval methods are exercised on the data produced by the simulator. Parameters of the X-band spaced-antenna design are then optimized. Benefiting from the parametric fitting procedure inherent in the time domain slope at zero lag and full correlation analysis, the study finds both of these algorithms to be more immune to thermal noise than the spectral retrieval algorithms investigated. With appropriately chosen baselines, these time domain algorithms are shown to perform adequately for 5-dB SNR and above. The study also shows that the Gaussian slope at zero lag (G-SZL) algorithm leads to more robust estimates over a wider range of beamwidths than the Gaussian full correlation analysis (G-FCA) algorithm. The predicted performance of the X-band array is compared to a similar spaced antenna implementation on the S-band National Weather Radar Testbed (NWRT). Since the X-band signal decorrelates more rapidly (relative to S band), the X-band array accumulates more independent samples, thereby obtaining lower retrieval uncertainty. However, the same rapid decorrelation also limits the maximum range of the X-band array, as the pulse rate must be sufficiently high to sample the cross-correlation function. It also limits the range of tolerable turbulence velocity within the resolution cell.
New far-field extrapolation method for the computation of electric fields