小波变换在面波插值中的应用
来源期刊：中南大学学报(自然科学版)2012年第11期
论文作者：周聪 刘江平 曾祥芝 范承余 曹进
文章页码：4395 - 4403
关键词：小波变换；地震道插值；瑞雷面波；频散能量；线性动校正
Key words：wavelet transform; trace interpolation; Rayleigh wave; dispersive energy; linear normal moveout
摘 要：利用瑞雷面波的频散信息可有效反演浅层横波速度结构。但在大时空采样地震记录中，由于瑞雷面波的速度很低，其记录中的高频成分容易出现严重假频，甚至出现超周期现象，这会降低其频率速度谱的信噪比，影响频散曲线的提取精度，在高频段更加如此。基于小波变换原理，实现对瑞雷面波的时空插值；根据瑞雷面波同相轴线性的特点，将线性动校正引入到插值中，改善高频信息的插值效果；进行理论模型数据和实际记录的试算与频率速度谱的提取以及插值前后频率速度谱的对比分析。研究结果表明：基于小波变换原理的插值方法是正确的；该方法可以有效提高频率速度谱的信噪比和频散曲线的提取精度，为大时空采样地震记录瑞雷面波频散曲线的提取提供理论基础。
Abstract: The dispersion information of Rayleigh wave can be used to calculate shallow S-wave velocity structure effectively. But in the large time-space sampling seismic data, the high-frequency signal components are usually aliased seriously, or even exceeding the cycle of the wave because of the low velocity of Rayleigh wave. This can decrease the SNR (signal-noise ratio) in frequency-velocity domain, and affect the accuracy of dispersion curves, especially in the high-frequency section. The interpolation of Rayleigh waves was accomplished based on wavelet transform (WT). According to the characteristic of linear seismic event, linear normal moveout (LNMO) was introduced into the process of the interpolation which could improve the interpolation effects of the high-frequency components. This method was used to calculate both synthetic and real data examples, extracted the frequency-velocity spectrums, and analyzed their frequency-velocity spectrums before and after interpolation. The results show that the interpolation method is correct, and it can improve the SNR in frequency-velocity domain and the accuracy of the dispersion curves. All these provide theoretical principle for extracting dispersion curves of the Rayleigh wave in the large time-space sampling seismic data.
周聪1，刘江平1，曾祥芝1，范承余2，曹进2
(1. 中国地质大学(武汉) 地球物理与空间信息学院，湖北 武汉，430074；
2. 安徽省高速公路控股集团有限公司，安徽 合肥，230051)
摘 要：利用瑞雷面波的频散信息可有效反演浅层横波速度结构。但在大时空采样地震记录中，由于瑞雷面波的速度很低，其记录中的高频成分容易出现严重假频，甚至出现超周期现象，这会降低其频率速度谱的信噪比，影响频散曲线的提取精度，在高频段更加如此。基于小波变换原理，实现对瑞雷面波的时空插值；根据瑞雷面波同相轴线性的特点，将线性动校正引入到插值中，改善高频信息的插值效果；进行理论模型数据和实际记录的试算与频率速度谱的提取以及插值前后频率速度谱的对比分析。研究结果表明：基于小波变换原理的插值方法是正确的；该方法可以有效提高频率速度谱的信噪比和频散曲线的提取精度，为大时空采样地震记录瑞雷面波频散曲线的提取提供理论基础。
关键词：小波变换；地震道插值；瑞雷面波；频散能量；线性动校正
ZHOU Cong1, LIU Jiang-ping1, ZENG Xiang-zhi1, FAN Cheng-yu2, CAO Jin2
(1. Institute of Geophysics and Geomatics, China University of Geosciences, Wuhan 430074, China;
2. Anhui Expressway Holding Corporation, Hefei 230051, China)
Abstract:The dispersion information of Rayleigh wave can be used to calculate shallow S-wave velocity structure effectively. But in the large time-space sampling seismic data, the high-frequency signal components are usually aliased seriously, or even exceeding the cycle of the wave because of the low velocity of Rayleigh wave. This can decrease the SNR (signal-noise ratio) in frequency-velocity domain, and affect the accuracy of dispersion curves, especially in the high-frequency section. The interpolation of Rayleigh waves was accomplished based on wavelet transform (WT). According to the characteristic of linear seismic event, linear normal moveout (LNMO) was introduced into the process of the interpolation which could improve the interpolation effects of the high-frequency components. This method was used to calculate both synthetic and real data examples, extracted the frequency-velocity spectrums, and analyzed their frequency-velocity spectrums before and after interpolation. The results show that the interpolation method is correct, and it can improve the SNR in frequency-velocity domain and the accuracy of the dispersion curves. All these provide theoretical principle for extracting dispersion curves of the Rayleigh wave in the large time-space sampling seismic data.
Key words:wavelet transform; trace interpolation; Rayleigh wave; dispersive energy; linear normal moveout
