Transcription of Preliminaryreference Earth model Adam M. Dziewonski’ and ...
1 PhysicsoftheEarthandPlanetaryInteriors,2 5(1981)297 356297 ElsevierScientificPublishingCompany,Amst erdam PrintedinTheNetherlandsPreliminaryrefere nceEarthmodel* DepartmentofGeologicalSciences,HarvardUn iversity,Cambridge,MA02138( )2 SeismologicalLaboratory,CaliforniaInstit uteofTechnology,Pasadena,CA91125( )(ReceivedDecember3,1980;acceptedforpubl icationDecember5,1980)Dziewonski, , , ,25:297 ,500summarytraveltimeobservations,100nor malmodeQvalues,massandmomentofinertiahav ebeeninvertedtoobtaintheradialdistributi onofelasticproperties,Qvaluesanddensityi ntheEarth , ~velocities,lowP~ ,PREM, ,andofEarthtidesandfreeEarthhavedevelope drapidlysince1950,muchoscillationsingeop hysics, ,forinstance,polarmo-Expansionintheutili zationofcomputersmadetion,Chandlerandann ualcomponents,anddi-itpossibletoconstruc tmanytypesofEarthmod-urnalmotionoftheEar th, ,emphasizesthevalueofthecontributionhadt woconsequences.
2 Ofseismologyfortheseresearches.(I)Therew asadifficultyofchoiceofanade-Itwasverydi fficulttosetupmodelsofthequateEarthmodel forresearchersthatdependonEarth sstructurebeforetheadventofcomputers;the structureoftheEarth,suchasthoselistedint hemoreimportantonesweresetupbyBullen,the firstparagraph.(2)Severalresearchersadop tedsomeproperties Anoteonthecalculationoftravelondmodel,wi ththeconsequencethattheirmodelstimesinat ransverselyisotropicEarthmodel byL~ (thisissue, ).Thesedifficultieswerepointedout,ataSym -298posiumonEarthTides, (Vicente, ,1973, , ),andinformalreferencemodelwerediscussed ,includingal-discussionsonthesubjectledt othesettingupofalowanceforattenuation;it wasdecidedtoinviteworkinggroup,composedo fmembersoflAGandcolleaguestoproduceandpr esentcompletemod-ISPEI,calledthe StandardEarthModelCommit-elsworkedoutbyt hemselvesandsatisfyingthetee ; ( ,Geo-upastandardmodelforthestructureofth eEarth, ,EOS,etc.)
3 Withtheannouncementthatfromthecentertoth esurface, (1977)wascon-dependedontheEarth ,correspondingtore-sub-committeesdealing withdifferentregionsofsearchesdonebyD, , , :constructmodelstakingaccountofdamping,t hat(I)thehydrostaticequilibriumproblem;( 2)theis, ;(3)theuppermantle;(4)regionD ;(5)coreconsideredthattheeffectsofattenu ationwereradius;(6)P-velocitydistributio ninthecore;andimportantandshouldbeconsid ered;but,sinceQ(7) ,insteadofhavingoneDuringthemeetingofthe Symposiumonreferencemodel,weshouldhavetw oreferenceMathematicalGeophysics(Banff,1 972),threere-models oneincludingQvalues andtheothersearchgroups, , ,presentedmodelsgivingtheAtthismeetingit wasevidentthattheoriginalmainparameters; inspiteofthefactthatthesereferencemodele nvisagedwasgrowingmoreandmodelsemployedd ifferentdatasetsandcomputa-moredetailed, thankstotherapidprogressoftiontechniques , , , (1973).
4 (1973) (Caracas,1978),withthestatementthatdecid edtocallthemodelfortheEarth sstructurethegrossEarthdataemployedforth econstructiona referencemodel ,followingtheexampleofofthemodelwasbeing enlarged, , (Canberra) , (1974). (Grenoble)itwasevidentthatitwouldbetoPhy sicsoftheEarthandPlanetaryInteriors, , ,wefollowtherecommendationsoftheSub-fit, suchcommentsshouldbeinthehandsofcommitte eonParainetenzation, , ,whichtheCorn-(1974).mitteehopestobeable toregardastheconclusionSection3containsa descriptionofthedatasetsofitswork, , ,Secretaryabriefoutlineoftheinversionmet hodused,theFortheStandardEarthMode~:n~m~ ~final(intermsofthisreportonly,ofcourse) , ,PREM, ( ) , ,however,tabulatedresultswaveandbodywave data,wehavefounditneces-forthe equivalent isotropicEarth.
5 Equivalent ,fortheuppermantle, ,notthatitprovidesanequivalent,orWeprese nttablescontainingthevelocities,elastics atisfactory, ,andarecalculatedwithaVoigtaveragingsche meand,auxiliaryparameterssuchasgravity,p ressure,therefore, , ,aconceptevaluatethemodelandthetablesand tocontrib-discussed,and,ingeneralterms, ,ReferenceEarthModel, ,inLima, ~seismicvelocities, ,oforderbe-AnaverageEarthmodel,thesubjec tofthistweenzeroandthree,todescribeseism icvelocities,work, :offreeoscillationsortraveltimesofbodywa ves,assumingthatoceaniccrustcoverstwo-th irdsofbuttherearenoequivalentexpressions fordif-theEarth , , (I) (2) 10s.
6 (3)Regionabovethelowvelocityzone(LID), (4)Lowvelocityzone(LVZ).(5) (6)Transitionzonespanningtheregionbetwee nTherearethreeprincipalsubsetsofdata:the 400and670kmdiscontinuities.(7) ,listedinthe(8) ,wereusedas(9) , :fromandthatthesameistruewithrespecttoat leastupto4Xl0~forsomeofthetoroidaloverto nestwomajordiscontinuities,althoughtheac tualtothe4X106recentlyreportedforthefund a-velocitygradientsarestillunresolved, , (1969),DziewonskiandGilbertisseparatedfr omtheremainingonesbyclearly(1972,1973),M endiguren(1973),Gilbertanddefineddiscont inuitiesthantoalteramodelthatDziewonski( 1975),andBulandetal.
7 (1979).l (1970),methodsofconstructionofsynthetics eismogramsDziewonski(1971),Dziewonskieta l.(1972),Millscansatisfactorilytreatabru ptdiscontinuities,but(1977),andRobertNor th(personalcommunica-arenotsuitableforre gionswithverysteepvelocitytion,1980). , , (1970),Dratlereta!.residualsforothertrav el-timestudies;thesetwo(1971),Sailor(197 8),SailorandDziewonski(1978),figureswill bediscussedmoreextensivelylateronSteinan dGeller(1978),Bulandetal.(1979), (1979),andSteinandDziewonski(1980).Allev entsanalyzedwereshallow, , ,werenotparticularlyrelevanttothisObserv ationsoftraveltimesofbodywavesarereport, wehavealsoderivedaveragetravel-timemostn umerous; (butforshalloweventsonly).
8 Therethebodywavedataoverthenormalmodeob- aresignificantandsystematicdifferencesbe tweenservationsisthat,becauseofshorterpe riods, ; . ;someofthecontroversiesoncurvesarevirtua llyidentical, ,differencesbetweentravelportantindefini ngtheregionsofthemantlewhichtimesforregi onsofdeepandshallowseismicitywedisc~ 6sand,whatisevenmoresurprising,Inorderto obtainarepresentativeglobalbody-thereisa systematicdifferenceof3 90 canprobablybeletinsoftheInternationalSei smologicalCentreforexplainedbyinterferen cewiththeSKSphase,years1964 ,oneretainsapproximately82.
9 26000eventswithreportsofnearly2000000P-I twouldappearthattheeventsforregionswithw avearrivaltimesand250000S-wavearrivaldee pseismicity(trenches) ,thereexiststheimportantgroupedonlyonone sideoftheevent, ,particu-manyexperiments, , , ,each30 10C1~ ci(aci-00ci8ciciciwci---2-ci0~D/Es~ ,~Es~\~/~%%%%%%%0ci~ ..,/4I IIIIIII020406080100 Epicentraldistance(degrees) (1968).DashedlineisfromHerrinetal.(1968) . (EngdahlandJohnson,1974)andScS S(Jordantravel-timeobservationsfromnatur alsourcesweandAnderson,1974), ,wetothemantletraveltimes, , :SKKS,SKS,PKP(AB)sourceandpathbiasbutweh avebeenabletofindandPKP(BC).)
10 , ,SKKS mostlyforcomparison,includedeepsourceP-w avedataarelikelytosufferfromasr/2phasesh iftdataofSenguptaandJulian(1976)andS-wav ewithrespecttotheSKSphase(ChoyandRichard s,dataofSengupta(1975),Halesetal.(1968), and1975),unlessSKKSisHilberttransformedb eforeGognaetal.(1981).Differentialtravel times,PcP cross-correlationwithSKS(orviceversa).30 3I~ IIIIII6 I$-Stxit4xH $,tx-~::~ot~9ox~I:~:/~~:~~s~*P~~ ?o /~++~%~S000-b~~:J~+~~ -~ +A++-++++-6-++-IIIIIIII20406080100 Epicentraldistance(degrees) +:fromGognaetat.(1981).X:fromHalesandRob erts(1970).Solidlineisfortheverticallypo larized(SV)shearwaveanddashedlineisforth ehorizontallypolarized(SH) ,(1970),corephasedatafortheAB, (unpublished)decreaseinS-velocitygradien tbelow600kmwasandPKiKP-PcPdifferentialtr aveltimesofEng-dictatedbytheneedtoobtain intersectionwithdahleta!
