Multi-domain Prestack Noise Suppression Technique Application Development Research

The ratio of signal and noise is the basis in seismic data processing.Highresolution and high fidelity couldn't realize without higher S/N ratio. And Theprocessing technique and method is useless without that. But there are kinds of noisein the seismic record.These noise will reduce the S/N ratio of seismic data.Prestacknoise elimination is a key measure to improve the S/N ratio and stack quality.Usually,different method of noise attenuation have different pertinence andadaptability.Without a method is valid to all kinds of noise.Therefore , theelimination of noise should be processed in multi-domain.In the paper,firstly,we have discussed the cause,the distributing,the degree ofcomplex and the type of kinds of noise and especially,have discussed thecharacteristic of these noise in detail.Secondly,we analyse the peculiarity differencebetween noise and signal in multi-domain and amplify their difference in adomain.Then ,we can design an arithmetic operator of noise suppression and realizethe separation of signal and noise. And these methods of noise suppression havebeen demonstrated according as the theory of noise generation and characteristic.Wechoose compatible domain and corresponding NA method to process.At the sametime ,we have studied the relevant processing module and software and realizedthem. Thirdly,we have chosen the representative sismic data which include typicalnoise from different area and processed them by using prestack noise suppressiontechnique in multi-domain and their combination.The effect of processing isobviously. And we reached the destination of improving S/N ratio and solution.Thisindicate that the method is feasible and effectual.It can improve the S/N ratio andresolution of seismic data. Finally,we have processed seismic data by the techniqueof prestack multi-domain noise elimination and their combination and archived nicerwhole effect .Then the processing flow of prestack multi-domain noise elimination isbuilt.By studying of prestack multi-domain noise elimination,we have concluded asfollow:①Zone abnormal noise suppression technique is a very good mothod ofprestack noise attenuation.The module of Zone Abnormal noise Processing(ZAP) ofOMEGA sps is an effective mothod of suppressing noise in Shot gather and CMPgather.②The noise of alternating current is suppressed by notching filter in theseismic data which dominating frequency is far from 50Hz.Otherwise,it issuppressed by mono-frequency noise suppression in the time domain of CGGsps;③The module of F-X Coherent Noise Suppression( FXCNS) of OMEGA sps isa good mothod to suppress coherent noise.We can use it when we knew the velocityof linear noise.And the mean weigh technique is another effective method ofcoherent noise suppression.We can try the method of T-X domain linear noiseattenuation of CGG sps.④The method of predict deconvolution of all seismic dataprocessing system can attenuate multiple wave.The transform of Radon is also amethod to suppress multiple wave.But the method is more complex than velocityscanning and inside muting.The method of inside muting is simple and effective andit adapt to higher coverages data.Its shortcoming is lost some information of near theoffset.The method of Adaptive Noise Cancellation Filter of OMEGA sps canestimate the multiple wave and it is an effective techniques.There is multiple waveattenuation method in F-K domain in the CGG sps.The best method to suppressmultiple wave is combined these methods;⑤In CGG sps ,there is random noiseattenuation method of FXNAT and FXYNA.And there is RNA and RNA₃Dcorrespond to those in OMEGA sps.But the effect of RNA and RNA₃D is moreobviously than FXNAT and FXYNA.We should pay attention to math the seismicdata after NA with no NA.