相关法从常规天气雷达信号中获取多普勒诺宽的模拟试验

A SIMULATION EXPERIMENT FOR EXTRACTING DOPPLER WIDTH FROM CONVENTIONAL WEATHER RADAR SIGNALS BY THE CORRELATION METHOD

理论上,对常规天气雷达信号进行相关分析可得多普勒谱宽,但此法是否有实用价值即实测精度如何是一个需要探讨的问题。用模拟方法对此进行研究是比较理想的既经济又灵活。本文模拟了一个可以改变谱形参量的接近实际的随机多普勒复频谱,对复频谱进行傅里叶反变换得到天气雷达回波的时域信号。然后对时域信号进行相关分析以求得自相关曲线和多普勒标准差。通过不同条件下多次试验结果的分析,得出不同的信噪比及多普勒谱形对相关法测量多普勒谱宽精度的影响,这对于用相关法从实际的常规天气雷达信号中获取多普勒谱宽有一定的指导意义。

Theoretically, Doppler width can be obtained from conventional weatherradar signals by correlation analysis. However, the applicability of this method or itsactual accuracy requires further proof. Simulation is an ideal, economical and quick wayto do this. In this study, a simulation is conducted of a random Doppler multi-frequen-cy spectrum that can change the spectrum parameters and the result is close to the actualconditions. The time domain signals are obtained from the weather radar echoes bymeans of Fourier inverse transform on the multi-frequency spectrum. Thereafter, acorrelation analysis is performed of the time domain signals, yielding the lag correlationcurve and Doppler standard deviation. Analysis of the experiment results under differentconditions reveals the effects of the different SNRs and the Doppler spectrum forms onthe accuracy of the correlation method, thus providing a guideline for the application ofthe method.