二维小波变换在去除面波干扰中的应用

Application of 2-D wavelet transformation in eliminating surface wave interference

(中国矿业大学资源与地球科学学院，江苏徐州221008)

College of Resources and Geoscience,China University of Mineral and Technology, Xuzhou 221008, China

在原始地震记录中，面波干扰往往会掩盖有效信号，降低资料的信噪比。常规的一维滤波和F K滤波方法在去除面波时，会损害有效信号，丢失一些有用的地质信息。小波变换是一种时频分析方法，具有分频和局部分析能力，可以从地震资料中将面波分离出来，而不伤害有效波。二维小波变换去除面波的基本原理是：首先依据面波和有效波视速度的范围，在时间 空间域把原始单炮记录分成没有面波和含有面波的记录；然后应用二维小波变换对含有面波的记录进行时频分析，将其变换到时间、频率、空间和波数四维域中，利用有效波与面波在频率域和波数域上的差异进行面波分离， 去除面波，提取有效波，并进行信号重构；最后将重构的信号与没有面波的记录叠加，即可得到去除面波干扰的资料。实际资料处理结果表明，二维小波变换去除面波的方法具有良好的应用效果，有效信号得到了加强，信噪比得到了提高。

In original seismic data, surface wave interference often conceals the effective signal and reduces the S/N ratio of data. The conventional 1-D filtering method and filtering method may damage effective signals and lose some useful geological information in eliminating surface wave. So the technique of using wavelet transformation to eliminate surface wave was discussed. It is a time-frequency analysis method which has the ability of frequencydivision and partial analysis. The method can separate surface wave from seismic data without doing any harm to the effective wave. The principles of this method are that 1) according to the distribution of the apparent velocity in surface wave and effective wave, the original seismic data are divided into nonsurface wave and surface wave in the timespace domain, 2) timefrequency analysis is made based on the data with surface wave through the 2-D wavelet transformation and the data is mapped into a 4-D domain related to time, frequency, space and wave number, 3) the difference between effective signal and surface wave in frequency field and wave number field are utilized to separate surface wave, remove surface, pick up effective wave and reconstitute signal, and 4) the reconstituted signal and non-surface-wave data are stacked into the data without surface wave interference. The actual data processing shows that the 2-D wavelet transformation in eliminating surface wave interference has good application results, the effective signals are strengthened, and the S/N ratio is improved.