浅层地震数据处理与去噪的应用

Application of Shallow Seismic Data Processing and Denoising

摘要: 在浅层地震勘探中有许多干扰波，其中面波最为常见。面波的特性为低频、低速、强能量。由于面波在浅反射地震记录中出现比重大，严重干扰所收集数据的信噪比。在后续的干扰波去除工作中因为面波繁杂的特性十分难以继续开展。因此，有必要找到一个可以压制和衰减面波的同时不损失有效信号同时提高信噪比的方法。瑞利面波是最低频率的低速次声波，其抑制方法就是基于这些特点。想要将二维地震数据较好的转换到频率–波数域，必须关联二维傅里叶变换来替代常规滤波器。采用二维离散小波变换设置独立参数的F-K滤波器，然后通过把数据转换到四维函数的压制过程。使用各种方法处理小波系数，并获得分解和重构。面波抑制后的浅表面反射率数据与原数据进行比较研究。处理的结果经过实际数据验证后有一定改善。将各个单道地震数据转化到时间域可以替代常规压制面波方法。在时间–尺度域中，地震数据中存在能量集中区，用设置好的衰减系数处理原数据，将所得到的压制后的数据和方法进行对比选优。

Abstract: There are many interference waves in shallow seismic exploration, among which surface waves are the most common. Surface waves are characterized by low frequency, low speed, and strong energy. Due to the large proportion of surface waves in the shallow reflection seismic records, the signal-to-noise ratio of the collected data is seriously disturbed. In the subsequent interference wave removal work, it is very difficult to continue because of the complex characteristics of surface waves. Therefore, it is necessary to find a method that can suppress and attenuate the surface wave without losing the effective signal and improve the signal-to-noise ratio. Rayleigh surface waves are low-speed infrasound waves with the lowest frequency, and their suppression methods are based on these characteristics. In order to better convert the two-dimensional seismic data to the frequency-wavenumber domain, a two-dimensional Fourier transform must be associated to replace the conventional filter. Two-dimensional discrete wavelet transform is used to set the F-K filter with independent parameters, and then the data is transformed into a four-dimensional function through the suppression process. The wavelet coefficients are processed using various methods, and decomposition and reconstruction are obtained. The shallow surface reflectance data after surface wave suppression is compared with the original data. The processing results have been improved to some extent after being verified by actual data. Converting individual single-channel seismic data to the time domain can replace conventional suppression surface wave methods. In the time-scale domain, there is an energy concentration area in the seismic data, the original data is processed with the set attenuation coefficient, and the obtained suppressed data and the method are compared and optimized.