The Research Of Prestack Space-time Gaussian Beam Depth Migration Based On Wavelet Reconstruction

With the continuous development of resource exploration,the exploration target gradually extends from the shallow layers to the deep and ultra-deep layers,and the resource types extend rapidly from conventional to unconventional.The exploration and development of deep reservoirs is of great strategic significance for the stable production and increasing reserves of old oil and gas fields.The essence of seismic imaging technology is the process of mapping the seismic data observed in the field to the imaging domain according to the propagation law of seismic waves by using mathematical and physical methods,so as to make the reflection wave locate and the diffraction wave converge,thus reconstructing the spatial distribution of geological structures.The exploration and development of deep reservoirs puts forward higher requirements for seismic data processing technology.One of the bottlenecks is how to achieve fast and high-precision imaging for deep reservoir targets.Seismic imaging technology is facing new challenges and thirsts for a more robust and reliable migration method.The rapid development of the current seismic migration imaging methods is inseparable from the innovation of computer technology and the improvement of basic imaging theory.The reverse time migration based on the two-way exact solution of wave equation is the most accurate migration method developed up to now,but it has some problems such as huge memory expenditure and low computational efficiency.Fourier finite difference migration can be applied on both the time domain and the frequency domain,but its imaging accuracy cannot be compared with the reverse time migration.The unparalleled advantage of the computational efficiency of Kirchhoff migration makes it widely used in massive real field data processing,but the imaging accuracy is low,which makes it unable to meet the needs of fine exploration.The Gaussian beam migration method not only inherits the flexibility and high efficiency of ray-based methods,but also has the matching ability of the irregular observation data.For complex high-steep structures and fault blocks,the imaging applicability of the Gaussian beam migration method is comparable to that of the wave equation migration.Aiming at the problems of fine exploration of deep reservoir,due to the large scale of the model and the difficulty of high-precision velocity modeling,the space-time Gaussian beam migration method has attracted much attention because of its advantages of flexible target processing,efficient calculation of data and weak dependence on initial velocity field in the depth domain.Its targetoriented imaging capability is expected to make it a mainstream seismic data processing technology for the fast and high-precision imaging research of deep reservoirs.In view of this,we select the space-time Gaussian beam migration method as the research object of this paper,which is of great significance to the practical production.The effective beam width of space-time Gaussian beam increases hyperbolically with the arc length of ray.When it propagates to the target imaging area,the plane wave front is no longer maintained,which results in poor imaging quality,inadequate illumination,and low calculation accuracy of paraxial ray travel time and amplitude in the target area.Considering the imaging accuracy,computational efficiency and imaging difficulties of deep reservoirs,this paper starts from the space-time ray theory and the basic theory of space-time Gaussian beam,constructs high-precision backward continuation wave field by using the up-going ray tracing strategy in model space,and derives the adaptive beam shape function driven by the initial velocity model and the target geological body.Thus,a set of efficient,robust theoretical method,optimization algorithm and practical technology module for complex structure conditions of the space-time Gaussian beam migration have been developed to serve the fast and highprecision imaging research of deep reservoirs.Specific research contents can be divided into the following steps:(1)Based on Gabor decomposition wavelet reconstruction theory,the space-time Gaussian beam forward modeling simulation method suitable for arbitrary wavelet is discussed;(2)By introducing up-going ray tracing strategy in model space,the high-precision back-propagation seismic wave field is constructed by using approximate time-domain Green function.And applying the cross-correlation imaging condition in a specific target time window,a space-time Gaussian beam migration method based on wavelet reconstruction is developed;(3)Based on the dynamic parameter control strategy,the adaptive beam shape function driven by the initial velocity model and the target geological body is deduced to optimize the propagation pattern of space-time Gaussian beam,and a space-time adaptive Gaussian beam migration method is proposed;(4)By introducing the inner product imaging condition of elastic reverse time migration,the elastic space-time Gaussian beam migration and space-time adaptive Gaussian beam migration for multi-wave and multi-component seismic data are developed.The stability,reliability and superiority of the above methods are verified by model trial calculation and actual data processing.