Seismic Diffraction Imaging Based On Path Integral

With the increasing demand for mineral resources and the decreasing reserves of large oil and gas fields easily exploitable in shallow layers,the oil and gas resources are shifting towards deep buried,complex and unconventional at present.Geological bodies such as faults,taper-outs,fissures,lenses,and cavities have gradually become the main targets of exploration,and the detection also has been asked higher accuracy.However,restricted by data acquisition and processing methods,conventional reflection seismic imaging technology is difficult to efficiently identify the above-mentioned small-scale irregular heterogeneous bodies(regions).Due to the lack of smooth and continuous wave impedance interface,these geological bodies are generally difficult to produce sufficient reflection response,but still they will cause rich diffraction carries high-resolution underground heterogeneous information.One can delineate heterogeneous bodies(regions)by accurately imaging the diffracted wave,thereby improving the accuracy and reliability of seismic exploration in complex areas.According to the classic theory of diffraction imaging,this thesis develops a post-stack seismic diffraction migration imaging method based on path integral,which can perform efficient and accurate imaging of zero-offset diffraction data.Firstly,solve the isotropic zero-offset velocity continuation partial differential equation to obtain its analytical solution in the f-k domain;Secondly,combine the path integral theory to construct a basic migration filter,and change its filtering characteristics by adjusting the integral limits;Finally,use the obtained filter to filter the post-stack diffraction data mapped to the f-k domain,and then convert the result back to the t-x domain to get the final imaging section.In addition,to further ameliorate the imaging effect,a weight factor can be added into the original filter.In principle,this method effectively highlights the diffraction source by using amplitude and phase control in the f-k domain to selectively increase the amplitude of the diffraction apex,and weaken the flanks with a certain inclination simultaneously.Because of the combination of path integral ideas,it does not need to provide a migration velocity model in advance,which reduces human influence and simplifies the imaging process.In this thesis,constant velocity continuation,unweighted path integral method and Gaussian-weighted path integral method are successively verified their feasibility through the zero-offset section of typical forward diffraction models,and the control effects of migration parameters on imaging are also studied.The results show that: constant velocity continuation is the migration basis of this thesis,and the diffraction events show different migration states in this algorithm under the deviation between the given migration velocity and the accurate one;Unweighted imaging method is equivalent to the continuous summation of all constant velocity continuation sections in a given velocity range.Diffracted waves' vertices stably superimpose in the process,and their steep wings offset by the continuous phase shifting;The Gaussian-weighted imaging method can be considered to use weight control to improve the superimposing contribution near the accurate velocity,meanwhile diminish the proportion of incorrect migration components in the final result.It can ameliorate the coordination of the imaging section.Furthermore,an imaging parameters selection principle of the path integral method is preliminarily discussed,then the actual seismic data processing is carried out based on this guide.Under the premise of diffraction's complete extraction,this method is capable of effectively imaging for heterogeneous bodies(regions),and its imaging resolution is better than conventional post-stack migration methods.