论文详情
:在复杂地表工区开展主动源地震勘探时,受崎岖地表以及近地表多尺度非均匀体等因素的影响,采集的地震数据中往往散射干扰波发育,采用常规的时频域滤波方法难以将其合理滤除。介绍了一种利用散射干扰波与有效反射波信号出射地表时的传播方向偏差,从角度域入手在叠前进行信号矢量分解滤波的地震数据去噪方法。数值模型实验结果和去噪前、后的实际单炮地震记录对比结果均显示散射干扰噪声和有效反射波信号能够被准确分离。我国西部某工区生产数据的处理结果表明,利用该方法滤除近地表散射干扰波后地震剖面成像质量得到了有效提升。当地震数据中噪声与有效信号在分布和时频特征等方面接近,但存在传播方向偏差时,该方法可以作为时频域、小波域等去噪方法的一种有效补充。
When active-source seismic exploration is carried out on complex near-surface areas,the acquired seismic data often include many interferential scattered waves.These scattered waves cannot be rationally filtered out via conventional time-frequency domain filtering methods.To remove them,an angle-domain vector resolution denoising method was introduced herein.This method exploits the differences in the emergence angle of the reflected and scattered waves and performs prestack vector decomposition filtering in the angle domain.Results of numerical and real single-shot seismic data processing demonstrate that the reflected and interferential waves could be separated satisfactorily.Field data processing from Western China reveal that the proposed method could improve the quality of the imaging profile.It is concluded that,if the noise and useful signals are similar in the time-frequency domain,but different in the directions of propagation,the method presented herein can play a complementary role to the denoising methods in time-frequency domain and wavelet domain.
国家重点研发计划项目(2018YFC1503200)、国家自然科学基金项目(41730425,41790463)、中国地震局地球物理研究所基本科研业务费专项(DQJB19B25)和中国地震局地震科技星火计划项目(XH18063Y)共同资助。