【摘要】 反褶积是地震数据提高分辨率处理的重要技术之一,零井源距VSP下行波反褶积被认为是较为理想的地震反褶积方法之一,但在存在较强干扰时,经过零井源距VSP下行波反褶积处理的数据与地面地震数据会经常出现较大的闭合差,以至于严重影响到零井源距VSP数据在地震层位标定和地质解释中的有效应用。实际零井源距VSP数据的研究发现,当零井源距VSP采集数据中存在较强的层间多次波和井壁耦合谐振时,经下行波反褶积处理后的VSP成像数据与地面地震数据之间出现了闭合问题。对影响闭合的因素进行了分析,在此基础上提出了上行波统计反褶积方法,实际资料处理试验表明,该方法可以较好地回避和压制以上干扰,获得更稳定的反褶积算子和反褶积效果。上行波统计反褶积有与地面地震反褶积相同的输入统计数据和算法,通过上行波统计反褶积处理,零井源距VSP与地面地震数据之间的闭合差问题得到了很好的解决。给出了基于上行波反褶积的零井源距VSP处理流程和质量监控方法。同时指出当VSP采集数据中含有较弱的干扰和层间多次波时,下行波反褶积仍是获得更精确反射系数的重要途径。更多还原

【Abstract】 Deconvolution is one of the most important technologies for improving seismic data resolution.Down-going wave deconvolution of zero-offset VSP based on is regarded as one of the ideal seismic deconvolution methods.However,when strong disturbance exists, zero-offset VSP data processed by down-going wave deconvolution has large misclosure with surface seismic data,which seriously impacts the effective application of zero offset VSP data in calibration and geological interpretation.Aiming at the problem,we carried out study.The results show that when zero offset VSP data existing strong interbed multiple and borehole coupling resonance has large impact on down-going wave deconvolution,and the VSP imaging data has misclosure with surface seismic data.We analyzed the factors impacting closure.Based on the research,we proposed up-going wave statistical deconvolution method.Actual data processing results show that the method can well dodge and press the disturbance,and obtain more stable deconvolution operator and deconvolution effect.Up-going wave statistical deconvolution has the same input statistical data and algorithm with surface seismic deconvolution. Through up-going wave statistical deconvolution processing, the closure between zero offset VSP and surface seismic data has been removed.The processing workflow and QC of zero offset VSP based on up-going wave deconvolution are given.Meanwhile, we pointed out that when weak interference wave and interbed multiple exists in VSP data,down-going deconvolution still is a important way for acquiring more accurate reflection coefficient.更多还原