Study On Quantitative Analysis Of Near-surface Seismic Wave Attenuation Characteristics

Because of the non-complete flexibility of formation medium,the seismic wave propagation is strongly affected by the formational absorption and attenuation,which causes the change of amplitude,frequency and phase of seismic wave.In order to accurately describe the propagation of seismic wave in the medium,it is necessary to consider the effect of absorption and attenuation.With the development and deepening of the seismic exploration,the near-surface geological condition is more and more complex.Compared with the deep formation,the near-surface absorption and attenuation has its own rules and characteristics.The quality factor Q is one of the parameters to evaluate the seismic wave absorb attenuation,which is the basis for the quantitative analysis of near-surface absorption and attenuation characteristics.Therefor,the study which develops the near-surface absorption attenuation characteristics and quality factor Q quantitative estimation has certain theory value and practical significan.Based on the seismic wave absorption and attenuation characteristics,the attenuation mechanism,attenuation theory(viscoelastic medium theory and two-phase medium theory)and the influence factors are summarized.The basic rules of the influence factors of the near-surface absorption and attenuation are clarified.Referring to other previous research,the near-surface absorption and attenuation characteristics,the estimation method of quality factor Q are summarized systematically,and the principles,the methods advantages and disadvantages are reviewed in detail.The estimation Q values generally show the sharp jumps,poor stability and lower accuracy because the influence of travel time error or velocity estimation error.Therefore,a new method estimating quality factor Q by the dual linear regression is given by learning from velocity regression analysis based on the uphole survey data.On the basis of the method principle and the algorithm precision theoretical analysis and compilling corresponding algorithm program based on the matlab platfoem realize the theoretical model construction,the synthetic uphole survey,the quality factor Q quantitative estimationThe results of the model test show that the error between the spectral ratio method and the bilinear regression method is within 3%.When the actual initial time is used to replace the theoretical time,the maximum error of the bilinear Q estimation is not more than 25%,and the average error is about 10%.The spectral ratio method maximum Q estimation error can be as high as 500%,the average error is about 92% when the reading error tends to zero.The noise is inevitable in the actual seismic record.So the random noise of the signal to noise ratio of 10 is added to the theoretical model.The bilinear regression method relative error is 35.4%,and the spectral ratio method is up to 100%.Although both errors are increasing,the accuracy and stability of the bilinear regression method are higher than the spectrum ratio method.On the basis of measured uphole survey date in Zhenba area,the experiments were carried out with the new method.The results of processing actual data based on the bilinear regression show that the spectral ratio method maximum of Q estimation of spectral ratio method is 3935.9,the minimum is about 9.02 and the difference is about 4000 under the influence of time reading error.In contrast,the new method can obtain more stable near-surface Q values,which are consistent with the near-surface velocity layers and have clear geological meaning,and it has good adaptability for different uphole survey system.