The Sparsing Constraint Registration Of PP And PS Waves In Multi-component Seismic Exploration

The benefits of multi-component data have been verified for imaging through gas clouds and for identifying shallow gas hazards.Multi-component seismic exploration makes use of both compressional(PP)and converted compressional to shear mode(PS)waves for improving seismic imaging and for extracting valuable additional information about subsurface physical parameters.A reliable joint interpretation depends on our capacity to identify and register PP and PS reflected wave events generated from the same reflectors.However,the registration is a crucial step in multi-component seismic data processing,since the traveltime is different for reflectors in PP images and the corresponding reflectors in PS images.Registration of PP and PS phases is a long standing problem in exploration seismology and a plethora of methods have been proposed for seismic data registration.In classic methods,we match the multi-component seismic waves by calibrating discrete time/depth axes manually.Several approaches have been developed to implement registration automatically.In this study,we propose an sparsity-promoting approach to solve multi-component registration problem.It establishes correspondence of reflection events in PP waves and PS waves data by estimating a warping function.The method proposed has been applied by Shengli oil field.The detailed research contents and the main results are as follows:(1)For multi-component registration problem,we propose a new curvelet-based registration method to improve the precision of registration,especially for the data with heavy random noises.By making registration in curvelet multi-scale spaces from coarser to finer scale,the proposed method is not sensitive to initial values of velocity ratio of PP waves and PS waves.Applications of the new method to real seismic dataset from Shengli Oilfield,China show good registered results in terms of both qualitative and quantitative analysis,in comparison with a traditional registration method and a wavelet-based method.(2)For further parameter study,we propose an L1-norm minimization approach to register the corresponding amplitude of reflection events in the PP-wave data and PS-wave data.The dynamic warping function is linearzed in the L2-norm fitting term,and a sparsity constrained is used to estimate smooth warping fun ctions.The latter is possible by defining the warping function in terms of a cosine basis.Sparsity is used to minimize the number of basis functions that are used to model the warping function.We have also applied a fast iterative soft thresholding algorithm for solving the non-linear registraition optimization problem.Numerical results of the linearized dynamic warping approach with L1-norm show that our method is better than traditional registration method and L2-norm method in terms of both qualitative and quantitative analysis(3)For the registration of sparse multi-component data,we propose a reconstruction-match method.First,we completed the interpolation based on the data-driven tight frame(DDTF)method.Then,we can get the results for sparse multi-component data by applying the L1-norm minimization approach.Numerical results show that our method is an effective method for muiti-component registration problem,especially when the seismic data has noise.