The Research On The Evaluation Approach And The Effect On Modeling Wavelet Extraction Method For High-Frequency Attenuation

Recently, the modeling wavelet extraction methods have been used widely in realseismic data processing. However it is based on the statistical hypothesis and the statisticalinformation of seismic data, the accuracy of which needs to be tested in real seismic dataprocessing. To the problem mentioned above, the thesis first summarizes the currentapproaches which can evaluate wavelet pick-up methods and analyzes the merits and defectsof them. Meanwhile the applicability of the approaches on modeling wavelet extraction andthe defects are discussed. Then the author has a deep analysis of seismic records attribution.The key attribution which has great influence on wavelet extraction is found, that ishigh-frequency attenuation on seismic records. Then we analyze and study the effect onwavelet extraction for high-frequency attenuation in seismic records and make an evaluation.At the end, the wavelet extraction method is improved against on the effect fromhigh-frequency and the accuracy can also be improved. Also it will provide a researchdirection for further improvement of the wavelet extraction approach.It is necessary to study the accuracy of modeling wavelet extraction methods applied inreal seismic data because of restriction by the assumption conditions. Recently, there are someapproaches for evaluation on wavelet extraction methods. However, these have somerestriction for modeling wavelet extraction methods. So in the thesis, we study the realseismic records attributes and the effects on wavelet extraction to evaluate the accuracy ofmodeling wavelet extraction method.Through the analysis of seismic records and wavelet attributes, high-frequencyattenuation is the significant characteristic of real seismic records which is not in conformitywith assumptions. And it has great influence on wavelet extraction. The absorption andattenuation of strata leads that high-frequency attenuation is prominent in real seismic records.Also the different seismic records have different frequency characteristic due to the differentblock and different processing of seismic records. Simulation experiments of synthetic records and real seismic data processing shows thathigh-frequency attenuation has great effect on modeling wavelet extraction method. We canfind the defect of the method through evaluating the wavelet extraction method, and improveit aim at the effect. Simulation experiments show that the improved method can effectuallyeliminate effect of high-frequency. The study also provides certain theoretical basis forexpanding the application of the wavelet extraction method and further improves the accuracyof it.