论文详情
复杂地表条件工区的地震资料往往存在严重的剩余静校正问题,常用的剩余静校正方法一般要求准确拾取初至旅行时,但在面对低信噪比地震资料时,很可能耗费极高的人力时间成本进行手动初至拾取。发展了基于高阶累积量时间延迟估计的自动剩余静校正方法,采用高阶累积量和相干叠加的方法得到两点间的时延,再使用时延的组合得到炮检点的剩余静校正量。该方法的优势在于,互相关使用两道数据进行时延估计,只利用二维地震信息,而高阶累积量使用多道数据,利用三维地震数据信息增强有效信号;与互相关类方法相比,高阶累积量不再要求噪声为高斯分布,与实际地震资料的情况更加吻合,对背景噪声的压制效果更优;结合地震干涉法中相干叠加的思路进一步提高信号有效叠加次数和信噪比,保证获得稳定可靠的道间时延估计结果。该方法采用完全数据驱动,在面对海量低信噪比地震资料时,可有效避免初至拾取所需的高昂时间人力成本。合成数据实例验证了高阶累积量进行时延估计的良好抗噪性及计算剩余静校正量的有效性。将该方法应用于中国西部某地区实际低信噪比地震资料的处理,明显改善了同相轴的连续性,与基于初至拾取的剩余静校正方法相比,该方法降低了初至拾取所需的大量时间人力成本。
Performing residual statics correction is challenging in areas with near-complex surface conditions.The conventional procedure requires accurate first-arrival information.For seismic data with low signal-to-noise ratio (SNR),however,this information is usually retrieved manually,making the method expensive and time-consuming.In this study,an automatic method was proposed,which entails a higher-order cumulant-based time delay estimation.The delay of the first arrivals was estimated using a higher-order cumulant (HOC) with coherent stacking.The residual statics at receiver and shot stations were subsequently calculated on the basis of the estimated delays.In the conventional procedure,a correlation-based method is implemented,which uses only two traces for delay estimation,and hence can only account for two-dimensional seismic information.In contrast,the HOC uses multiple traces,and can make full use of three-dimensional seismic data,thereby enhancing the effective signal.Moreover,differently from the correlation-based methods,the HOC does not assume a Gaussian distribution for the noise,and is therefore able to suppress it more effectively.The proposed method also implements coherent stacking,which is commonly used in seismic interferometry,to improve the SNR and thus obtain stable and reliable time delay estimations.As the method is completely data-driven,labor cost and time consumption are greatly reduced.Tests on synthetic data demonstrated the improved anti-noise performances of the cumulant-based method to estimate time delays,and also verified the effectiveness of the method for residual statics analysis.An application to actual seismic data with low-SNR in western China also showed a
国家重点研发计划项目(2018YFA0702503)和国家自然科学基金项目(41674122)共同资助。