Seismic Data High-precision Surface Multiple Separation Method

Surface-related multiple are widely present in oil and gas exploration and development,affecting the accuracy of seismic data processing and presenting new challenges to traditional seismic exploration techniques.In order to improve the signal-to-noise ratio of seismic data for multiple in seismic data,this paper focuses on surface-related multiple wave processing methods,mainly includes multiple suppression method and multiple separation method.The conventional multiple suppression method is mainly divided into two types: the filter method uses the kinematic difference between multiples and primaries to suppress multiples,and the predictive subtractive method is based on the wave equation,Multiples are suppressed using the dynamic difference between primary and multiple.The advantage of the predictive subtraction method is that it does not require prior knowledge of the subsurface media and can better adapt to complex areas.This method is based on the principle of the wave equation.It can suppress multiple and does not damage the primary.It is widely used in the oil exploration industry.In this paper,a variety of multiple suppression methods are introduced.The principle and characteristics of the surface-related multiple suppression method based on the L2 norm are described.Starting from the original data,multiple channels of seismic data such as frequency expansion are derived.According to the principle of least squares,the adaptive filter is obtained,and the primary data after multiple suppression is further obtained.The SRME method based on the wave equation predicts surface multiples without prior knowledge of subsurface media,and can be well adapted to areas with complex geological structures.However,the wavelet effect is generated during the multiples prediction process,resulting in the multiple frequency band prediction narrower than the actual multiple frequency band.According to the high-precision multiple surface wave separation method,the excess wavelet effect in multiple wave prediction can be eliminated,and the frequency band of seismic data can be broadened.This makes it possible to predict the travel times of multiple waves,and the amplitude and phase are in agreement with the actual multiples,resulting in good multiple suppression effects,provides data support for multiple imaging.This paper focuses on the principles and steps of the above method.Tests on horizontal layered model and SMAART model show that the surface multiple suppression method based on wave equation can effectively suppress multiples and improve the signal-to-noise ratio of seismic data ratio.By testing the SMAART model,the multi-frequency suppression method based on frequency expansion can also obtain good multiples suppression effect under strong noise.For the case of a more complex geological structure,this method is more effective than the parabolic Radon transform filter method in suppressing multiple waves and does not damage the effective wave energy.For the spectrum analysis of the horizontal layered model test results,the multiples obtained by the high-precision surface multiplex separation method are wider than the predicted multiples and the high-frequency energy is enhanced.