The Study On High-resolution Velocity Modeling And Reverse-time Source Imaging Methods In Passive Source Location

With the increasing requirements for oil and mineral resources in the national economic construction,the focus of mining and oil exploration has gradually shifted to the deep parts,and efforts have been made to identify the mineral resources in the second depth space,which puts forward higher requirements for the resolution of passive seismic location.Since passive seismic location mostly uses reverse-time wave field extrapolation and travel time based source location inversion method,in this way,we first need to obtain the accurate velocity parameters,model structure and lithology distribution in this region.To solve this problem,we first use the active source full waveform inversion method to obtain the velocity parameters.The FWI method makes full use of the kinematics and dynamics information of seismic waves in the process of restoring the subsurface velocity model.Therefore,the FWI method has its advantages of high resolution imaging,and can be well adapted to the requirements of high-resolution source location.However,while the FWI method has the advantage of high resolution inversion results,there are still many problems to be solved.Aiming at the strong nonlinear problem of the FWI objective function of seismic exploration,firstly,the inversion method of the normalized energy spectrum objective function is proposed.And then,based on the problem of the inversion effect of the strong scattering velocity model such as metal ore the joint inversion method of energy spectrum and envelope objective function is proposed.After that,the FWI results were used as input the velocity models for the inverse travel time source location algorithm,which can effectively reduce the location artifacts and errors caused by the inaccuracy of the velocity model.The three new methods and corresponding research results presented in this paper are briefly summarized as follows:(1)Starting from alleviating the nonlinearity of the FWI objective function,theinformation of the frequency domain energy spectrum is used to construct the misfit function of the seismic inversion,so as to alleviate the cycle skipping problem during in the waveform matching process.When the phase information of the synthetic data differs from the phase information of the observed data by more than half a cycle,a significant cycle skipping phenomenon occurs,and the objective function based on the energy spectrum information of the frequency domain can be effectively mitigated,in this way,it helps to provide a better initial velocity model for conventional FWI methods.At the same time,the FWI objective function based on the energy spectrum information also has the independent ability on the source wavelet and attenuating the noise component in the test data.(2)Based on the direct envelope inversion method of joint amplitude and phase information,a joint inversion method of energy spectrum and envelope objective function is proposed,which effectively weakens the dependence of seismic inversion method on the strong scattering velocity model such as the metal ore velocity model.While breaking the Born approximation,the PADEI and NESFWI objective functions are combined to give full play to the advantages of PADEI's macro-scale construction and the weak nonlinearity of NESFWI.Finally,the high-precision inversion results of the metal ore velocity model are obtained.In addition,a weighting factor is added into the hybrid objective function to control the proportion of the PADEI and NESFWI method gradient,and then the inversion accuracy of the inverted metal ore velocity model could be controlled.(3)The reverse time source location method mainly uses the reverse travel time imaging method to take the travel time field as the data source.In the process of source location,it is not necessary to extrapolate the wave field into the model space.This calculation efficiency is much higher than that of the conventional wave field based source location method.Traveltime source imaging method which is based on extreme value difference,in this case,we analyzes the imaging process of reverse travel imaging method in detail and the difference between the source and non-source location,in different ways.Based on it,a variety of new imaging conditions were proposed,and the real-time velocity model and inversion velocity model of the metal ore velocity were used to test effect of the source imaging conditions.In addition,to solve the dependence of the inverse travel time imaging method on the initial velocity model,we propose a velocity model construction method based on the classes of theFWI method to provide high-resolution velocity model as a input velocity model for the reverse travel time source location,and then realize the FWI based high-resolution reverse travel time source location method.