山区复杂构造低信噪比地震资料的修饰性处理

叠前去噪后，初叠剖面上还有随机、相干噪声，以及处理中产生的噪声，因此需要做叠后去噪，即修饰性处理。常用方法有相干加强，模式化加强，随机噪声衰减和去规则干扰，其原理是基于多道相干去噪。去噪具两重性，可提高信噪比，但会造成波形和构造形态失真。针对山区复杂构造低信噪比资料，提出了做好修饰性处理的思路和方法：①了解复杂构造顶部、翼部和地层变化的反射特征；②熟悉去噪方法的性能，及其参数对剖面特征的影响；③用压制干扰波来提高信噪比，有利于有效波的保真，采用f—x域滤波与f—k滤波相结合的方法，可降低对弯曲同相轴的影响程度；④解释与处理相结合，有利于压制规则干扰波和突出有效波。

Modified processing on low signal-to-noise ratio seismic data from mountainous area of complex structure

After prestack noise elimination, noises may still exist in the stacked data, so poststack noise elimination (modified processing) is sometimes nee- ded. The commonly used methods include coherent enhancement, pattern enhancement, random noise attenuation, and regular interference elimination. The theory behind these methods is the coherency of multiple traces. Denoising improves the signal-to- noise ratio (S/N). On the other hand, it causes dis- tortions of wave shape and structural features. This paper presents a procedure of modified processing suitable for low S/N data from mountainous areas. Firstly the reflection characters at the top and flank of complex structures are determined. Processors should have knowledge of the capability of different denoising methods and their influences on the seismic data. Filtering in both f--x and f--k domains re- duces the influence of the processing to curved e- vents. Combining interpretation with data processing is in favor of suppressing the regular noises and makes the effective wave prominent.