Research On Directional Seismic Illumination With Rugged Topography And Random Noise Suppression

No matter oil-gas exploration or mineral exploration is moving towards the zones where geological conditions are often complex and undulate,and the ore body often presents the heterogeneous nature.The natural and cultural environment of these places are generally complex.The seismic exploration based on the horizontal surface has difficulty to satisfy the requirement of exploration in the complex zones.It is difficult to receive enough reflection seismogram.Besides,geometry of the observation system and propagation of seismic wave in overlying structures both can seriously affect the quality of seismic data.What is more,except for the acoustic signals and plain wave,there is random noise in the seismic records from exploration zones with rugged topography.The random noise is difficult to predict and sometimes it can lead to distortion of desired signals.So the seismic data can be seriously affected by the random noise,bringing a huge challenge to seismic data processing.In consideration of rugged topography,we study the directional seismic illumination and random noise suppression,supporting for oil-gas exploration and mineral exploration theoretically.The seismic illumination is based on the exact numerical simulation of seismic wave.It is an important method to design seismic observation system and study distribution of seismic wave energy in the complex zones.Source array can excite directional seismic wave and illuminate subsurface structures.The previous researches are generally based on the horizontal velocity models.Practically,there exists too many shadows when illuminating complex zones to meet high requirement about signal-to-noise ratio.In order to compute seismic illumination when the surface is rugged,based on the previous researches and the source array theory,a method of directional seismic illumination on rugged surface is proposed.This paper employs varying grid finite difference method to simulate seismic wave in the model with rugged surface.According to the relief surface and the required direction of illumination,the horizontal coordinates are rotated into inclined coordinates.It makes it convenient to calculate the source array's excitation delay.The numerical examples on the models with rugged surface indicate that this new method is flexible and practical for computing directional wave field and illumination in areas with complex surface.Besides,this paper analyzes the influence of source spacing,source number and position of geophone array on the seismic date,which can guide seismic data acquisition system in the complex exploration zone.It's impossible to get seismic data without any noise.In order to improve signal to noise ratio of the seismic data,this paper studies seismic random noise suppression.Based on the sparse and redundant representation theory,we improve both the dictionary update stage and the sparse coding stage of the conventional dictionary learning algorithm.While keeping the supports intact,the dictionary atoms are recurrently updated to adapt to the specific seismic data.In the dictionary domain,large coefficients represent effective signals.Taking full advantage of this characteristic,several large coefficients from the last round of iteration are used as initial coefficients.By this way,the computational efficiency of the learning algorithm can be improved.The new algorithm is applied to synthetic seismic records and real data.Compared with the conventional K-SVD algorithm,the denoising results are satisfied.It indicates that the new method can remove the random noise and protect the effective information at the same time.It is competitive in improving the signal-to-noise ratio of seismic records.