The Influence Of Missing Seismic Data In Intenral Multiples Prediction And Its Soluiton

Multiples problems have always been an important subject in seismic dataprocessing. The presence of multiples interfere people with identifying primary waves,affecting the reality and reliability of seismic imaging, which makes the amplitude,frequency and phase of the reflection wave of target layer distort. Resulting in seismicinterpretation workers making wrong judgment, and may cause severe economic lossand increase the risk of oil and gas exploration.In the practical application, filtering based method for multiples attenuation isconfined more or less, and easily harm the primary wave. Wave equation basedtechnique does not require any known geological information, which has become thefocuses of the current seismic exploration research. Multiples are divided into surfacemultiples and internal multiples. The research for the former one has graduallymatured, the latter one due to weak energy, has been overlooked, but with the increaseof exploration degree, internal multiples issues are raised gradually, because it doesnot have the obvious characteristic and causes the depth part of the section existserious frequency dispersion, so it is very difficult to identify and remove them.In this paper, by using construct virtual events method to predict internalmultiples, using self-adaption iterative subtraction process to subtract it from theoriginal record, realizing the purpose of multiples attenuation. This algorithm belongsto the wave equation technique, with excellent amplitude fidelity, but at the same timeit has strict requirements to the seismic records, seismic trace missing, especially lack of near offset will seriously affect the prediction results, so the interpolation ofmissing seismic data become the premise of internal multiples attenuation based onconstruct virtual events.Due to seismic acquisition, observation system is very difficult to record thecomplete seismic wave fields, result in seismic data were often irregular and sparse,missing data not only lose some information, more importantly is to produceunnecessary noise in a variety of multi-channel process, so the reconstruction ofmissing data has become an important process in seismic data processing. Along withthe compressive sensing concept put forward in the field of signal processing,combining mathematical transformation and achieved good results in the interpolationproblem in seismic data processing. In this paper, using the characteristics of seismicsignals, as the continuous phase axis, complete seismic data interpolation by signalreconstruction technique. Seislet transformation is based on the formation of localcontact, the valid signal of adjacent seismic data has good correlation, and fullyconsidered the structural feature of spatial direction of seismic data, its application inthe missing data reconstruction had been fully displayed. Using the prediction errorfilter scaling invariance could effectively carry out anti-alias interpolation of seismicdata.Through the theoretical model test shows that building up virtual events methodto predict and attenuate internal multiples, and achieve good effects, simulate theaffection of the method caused by nearly offset and random missing traces, and thetheoretical analysis and experimental results will be given. By using seislettransforming frame of seismic data direction, through the calculation of the spatialaliasing seismic angle model, developing seislet transform anti-alias form to solve theinterpolation problem of seismic data. The results show that, seislet transformiteration method achieved more accurate interpolation to the missing seismic data,especially to the random missing traces, and provide complete data information forpredicting and attenuating internal multiples.