基于小波的地闪首次回击辐射场的多重分形分析

Wavelet-based multifractal analysis of the radiation field of first return stroke in cloud-to-ground discharge

利用2002年夏季青海野外观测慢天线电场变化仪资料,应用小波变换模极大方法对地闪首次回击辐射场的检测及多重分形谱特征进行了分析.发现小波分解在小尺度上的时间变异系数在回击主脉冲峰处(回击点)呈现明显的尖峰,据此可对回击点进行快速可靠的检测;地闪回击辐射场多重分形谱可用推广的多重分形二项倍增串级公式比较精确地拟合,最小标度指数及谱宽度平均分别为-0.11和1.5,是重要的回击特征参数,而小波及多重分形应是闪电信号处理的重要工具.

By using electric field change data obtained during the summer of 2002 in Qinghai province, we analyze the detection of the radiation field of first return stroke involved in negative cloud-to-ground lightning discharges and its muhifractal behavior with the Wavelet Transform Modulus Maxima （WTMM） method. We find that the variation of wavelet coefficients at the lower level of decomposition exhibit a sharp peak just at the position of the main pulse peak of return stroke （return stroke point）, thus the return stoke points can be reliably and effectively detected. We also find that the radiation field of first return stroke exhibits strong multifractality, and all scaling behavior can be fitted with a modified version of the binomial multiplicative cascade multifractal model. The two characteristic parameters, the mean width of spectrum and the minimum spectrum exponent, are 1.5 and - 0.11 respectively. We suggest that the wavelet and muhifractal method is a powerful tool for detection and reorganization of lightning signals.