Study On The Analysis And Attenuation Of The Multiple From Loess Region

Multiples disturbance is a common interference in seismic data, because of the thick loess and the big difference of wave impedance between the loess and the bedrock, the multiples are rich here. The existence of multiple leads to suppress the effective wave energy and reduce the signal-to-noise ratio, and then will mix with the effective wave, if one can’t recognize it and will make a wrong geological inference. The identification and suppression of multiples is a key link in seismic data processing, so identification and attenuation of the multiples has very important application value.In this paper, starting from the mechanism and classification of multiples, through setting up the forward model and analyzing the single shot record gathered by seismic exploration on the surface and VSP, learned multiples are rich in loess region and that is one of main reasons for poor reception. So depressing multiples is very necessary. Then introduce a variety of ways to identify multiple wave and several pop methods, such as predictive predictive deconvolution, F-K filtering, Radon transform and SRME and so on. Introduces the principle, algorithm, which is the basis of the various methods to analyses the applicable condition, the effect of suppressing and advantages and disadvantages and so on. Then make a forward modeling of geological model based on the information of the loess tableland to analyze the field characteristics of multiple wave, then recognize the model multiple wave in different ways. In view of the complex multiple, propose a combined method to depress the multiples. And designed a scheme combined predictive deconvolution, F-K filtering and Radon transform to remove the multiples.For the simulate data of the forward model, designed a combined method to remove the multiples and get a satisfying results and so do the real data, proved the feasibility of the method.