基于最小二乘偏移的地震分辨率分析

Seismic resolution analysis based on least squares migration

最小二乘偏移将地震成像过程视为一种最小二乘线性反演问题,能有效降低由数据不规则引起的偏移假象,改善窄频率或窄波数带宽、复杂变化子波和强吸收衰减所造成的偏移噪声,进而提高复杂勘探目标的成像分辨率,得到更接近真实地层反射系数的振幅信息,是解决深层油藏描述和油藏监测问题的重要手段之一.理论上,最小二乘偏移能完全消除采集的有限孔径和不规则性对成像的影响,获得最佳的空间分辨率和振幅精度,但在实际应用中,由于反演精度和迭代次数的限制,最小二乘偏移过程仍会受到采集因素的制约,需要对其效果进行评价.但另一方面,最小二乘偏移的分辨率特征及其采集需求与传统偏移具有明显的差异,传统的分辨率评估手段也不再适用.基于此,本文提出了一种基于最小二乘偏移的地震采集观测系统聚焦分辨率定量评价方法,能在采集施工前,针对复杂三维速度模型中的目标点、目的层或目标区域,基于最小二乘偏移算法定量预测待选采集方案的分辨率和振幅精度,进而反馈采集参数设计,从数据采集的源头提高复杂地区的地震成像效果.实例应用结果表明,在较复杂介质条件地区或存在地表采集障碍区的情况下,最小二乘聚焦分析得到的偏移噪声、分辨率和角度域振幅精度均要明显地优于一步法聚焦分析的结果,采用最小二乘聚焦分析可以实现更准确的采集评价.

The least squares migration considers the seismic imaging process as a least squares linear inversion problem, which can reduce the imaging artifact caused by data irregularity, and migration noises produced by the narrow frequency or wave-number bandwidth, spatial-change wavelet and strong absorption attenuation. The least squares migration can improve the imaging resolution of complex exploration targets, and obtains accurate amplitudes closer to reflection coefficients, which can be seen as a valuable tool to solve the problems of reservoir description and reservoir monitoring. In theory, the least squares migration can completely eliminate the negative influence of the finite-aperture and irregular acquisition on the imaging and achieve the finest spatial resolution and the most accurate amplitude. However, in practice, due to the limitation of the precision and number of iterative inversion, the least squares migration still suffers from incomplete and reregulate acquisition geometry, which needs to be evaluated before its implement. On the other hand, the least squares migration may result in different resolutions and amplitudes from traditional migration methods, but which cannot be evaluated by traditional resolution analysis methods. Based on this, we propose a quantitative resolution analysis method for seismic acquisition geometries based on least squares migration, which can quantitatively predict the resolution and the amplitude effect of candidate acquisition schemes for a target point, layer or area in a complex 3 D velocity model before acquisition implement. Using this method, we can optimize the acquisition parameter to improve the seismic data quality and thus improve seismic imaging in complex areas. Results show, in the case of complex medium or with acquisition obstacles,the least squares resolution analysis can provide the more accurate resolution and amplitude than the traditional resolution analysis.