Simultaneous seismic data denoising and reconstruction via multichannel singular spectrum analysis

《Simultaneous seismic data denoising and reconstruction via multichannel singular spectrum analysis》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、GEOPHYSICS,VOL.76,NO.3(MAY-JUNE2011);P.V25–V32,8FIGS.10.1190/1.3552706Simultaneousseismicdatadenoisingandreconstructionviamultichannelsingularspectrumanalysis11VicenteOropezaandMauricioSacchiintodifferentdomainsbyusingmethodssuchastheFourierABSTRACTtransform(

2、Sacchietal.,1998;Duijndametal.,1999;LiuandSacchi,2004),theRadontransform(Darche,1990;Tradetal.,Wepresentarankreductionalgorithmthatpermitssi-2002),andthecurvelettransform(HerrmannandHennenfent,multaneousreconstructionandrandomnoiseattenuationof2008).Anotherca

3、tegoryofmethodscaninterpolateregularlyseismicrecords.Webasedourtechniqueonmultichannelsampledtracesusingspatialpredictionfilters(Spitz,1991;singularspectrumanalysis(MSSA).ThetechniqueentailsAbmaandClaerbout,1995;Porsani,1999).Anextensionofpre-organizingspatial

4、dataatagiventemporalfrequencyintoadictionfilterinterpolationtoanirregulardistributionoftracesblockHankelmatrixthatinidealconditionsisamatrixofhasbeenproposedbyNaghizadehandSacchi(2007,2010).Inrankk,wherekisthenumberofplanewavesinthewin-general,mostseismicdatar

5、econstructionmethodsdonotdowofanalysis.Additivenoiseandmissingsampleswilladdresstheinterpolation/reconstructionprobleminthepresenceincreasetherankoftheblockHankelmatrixofthedata.ofnoise.Infact,denoisingandreconstructionareoftenconsid-Consequently,rankreductio

6、nisproposedasameanstoereddifferentproblemsandtreatedwithadiversearsenalofattenuatenoiseandrecovermissingtraces.Wepresentantools.iterativealgorithmthatresemblesseismicdatareconstruc-Theobjectiveofthispaperistopresentmultichannelsingulartionwiththemethodofproje

7、ctionontoconvexsets.Inspectrumanalysis(MSSA)asatoolforsimultaneouslydenois-addition,weproposetoadoptarandomizedsingularvalueingandreconstructingseismicdata.MSSAisadata-drivendecompositiontoacceleratetherankreductionstageofthetechniquethatarisesfromresearchona

8、lternativetoolsforthealgorithm.WeapplyMSSAreconstructiontosyntheticanalysisofmultichanneltimeseries.MSSAisanextensionofexamplesandafielddataset.Syntheticexamplesweresingularspectrumanalysi