Font Size: a A A

Method Of Wave Field Extrapolation For Multiples Elimination And Its Improvement

Posted on:2016-10-01Degree:MasterType:Thesis
Country:ChinaCandidate:H L SunFull Text:PDF
GTID:2180330467997500Subject:Earth Exploration and Information Technology
Abstract/Summary:PDF Full Text Request
Multiples attenuation is an important subject in seismic data processing[1],multiples removing effect will directly affects the back of the data interpretationprocessing, bringing certain difficulties for seismic data interpretation. An obviousexample is that the important geological structure will be shielded and cannot berevealed well. It may also lead to the error of deposit thickening and affect the wholeeffect of seismic exploration. In order to improve the effect of seismic exploration,repressing and identify multiples is a very important problem. The way to identify andsuppress multiples is to find the difference between multiples and the primary wave.There are many kinds of multiples, the classification methods also have a lot of kinds,according to the position producing multiples, the surface related multiple wave andintrabed multiples. As the name implies, the surface related multiples’ generationlocation of multiples is on the surface. In offshore exploration, the reflective surfacefor the surface of the water. The multiples’ reflection energy is strong. Interbedmultiples refers to the multiples which repeatedly reflect in underground layers. Inaddition to the classification method above, multiples can also be classified accordingto the length of the cycle, according to this classification method,multiples can bedivided into multiple wave of long period and short cycle[2].The characteristics oflong period multiples has a separate phase axis, with deterministic performancecharacteristics and can be recognized easily in the seismic data. In contrast, the shortcycle multiples’ causing reason is more about the underground medium thin layerinterference, and compared to the long period multiples, the short cycle multiples isextremely hard to identify[2]. Due to the importance of the problem of multiples, thusit has caused many researchers’ attention in the field of seismic exploration, a largenumber of research strength into the problem of multiples study. Many exploration seismologists work on multiples suppression, and acquire a lot of achievements. TheSEG annual meeting in1997, launched a discussion on suppressing multiples problem,and then, in1999, THE famous science and technology magazine "THE LEADINGEDGE opened up a special project on suppressing multiples to publish, reflecting thatthe exploration geophysicists had attention to this problem[3]. A lot of effectivemultiples suppression methods are constantly being displayed and get great promotion,becoming the indispensable steps in the process of seismic exploration. Thesemethods are divided into two kinds. One kind is based on the difference betweenprimary wave and multiples in space, another kind is based on the predictability andcyclical of multiples. Primary wave and multiples have difference in undergroundtransmission path. This nature make them have difference in speed and reflectionstructure. The first kind of elimination method is based on this principle. There isinner link between primary wave and multiples.We can use the explicit or implicitwave equation to link them. This method first need to predict multiples, then usingprimary wave, finally subtracting multiples from the original data. In these two kindsof methods, there are some common methods include the method based on time andangle difference, predictive deconvolution method, the wave field continuationmethod, SRME method and so on. The SRME method is to use prestack data to buildthe surface related multiples wave. Its basic process is to use prestack data itself as afactor of wave field continuation, namely through the convolution of the data itself toproduce the first-order multiple wave by primary wave. This is a data-driven method.The advantage of this method is not need to know the information of the structure ofthe underground media. Because the using data itself as continuation operator, whichalready contains the information about the structure of the underground medium. Butthe problems existing in the SRME method is that it is difficult to meet therequirements of the theory of actual sampling data, particularly in the field of3d.Aiming at the problem of SRME method, this paper propose the wave fieldcontinuation method bornning of huygens’ principle, making seismic wave fieldreceived on the surface extrapolated to the bottom of the sea. Then we extrapolate the result to the surface, namely adding a communication process in the water, increasingthe multiples’ order. Because this kind of method’s prediction operator is based on themodel, which is not limited to the sample, making up the flaw of SRME. Bycombining Curvelet domain matching subtractive method, the technology of wavefield extrapolation obtains the good multiples suppression effect. Curvelet transformas a new kind of mult—scale direction time-frequency analysis method, has verygood sparsity and direction. Its sparsity make its use few coefficient to accuratelydescribe the seismic signal. The direction allows the Curvelet transform to representthe edge of image more sparsely, and express the seismic signal efficiently, stably andoptimally. Curvelet transform’s advantage lies in that the effective signal and randomnoise are stored separately in the Curvelet domain. So we can extract the effectivesignals, at the same time, it can not enhance the random noise and interference.Curvelet transform’s anisotropic properties make it describe the seismic signal inthe optimal degree; Curvelet transform’s direction makes it be able to deal withseismic signal elaborately, namely dealing with the specific location of the seismicsignal, ignoring the rest of the signal. This provides a new method for the highresolution seismic data processing. Effective seismic signals and random disturbanceis very different after Curvelet transform, when processed, we only need to extract theeffective seismic signal, and then remove the interference of random, in order tosuppress random noise, improving signal to noise ratio.Even though multiples elimination is important, what we can not ignore is thatboth multiples and primary wave are reflected to the surface through the undergroundmedium,received in record. So both multiples and primary wave contain a largenumber of information of underground medium. It provides a new direction tosuppress multiples. We not only remove multiples, but also take use of thecharacteristics of multiples, improving the resolving power of underground medium.
Keywords/Search Tags:Curvelet Transform, seismic wavelet, Green function, wave fieldextrapolation
PDF Full Text Request
Related items