North Anatolian Fault

14 Pages·2017·1.6 MB·English

Fault interactions in the Sea of Marmara pull-apart
(North Anatolian Fault): earthquake clustering and
propagating earthquake sequences
Nicolas Pondard, Rolando Armijo, Geoffrey C. P. King, Bertrand Meyer,
Frédéric Flerit
To cite this version:
Nicolas Pondard, Rolando Armijo, Geoffrey C. P. King, Bertrand Meyer, Frédéric Flerit. Fault in-
teractions in the Sea of Marmara pull-apart (North Anatolian Fault): earthquake clustering and
propagating earthquake sequences. Geophysical Journal International, 2007, 171 (3), pp.1185-1197.
￿10.1111/j.1365-246X.2007.03580.x￿. ￿hal-00314342￿
HAL Id: hal-00314342
https://hal.science/hal-00314342
Submitted on 16 Jun 2017
HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est
archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
Geophys.J.Int. (2007)171,1185–1197 doi:10.1111/j.1365-246X.2007.03580.x
Fault interactions in the Sea of Marmara pull-apart (North Anatolian
Fault): earthquake clustering and propagating earthquake sequences
Nicolas Pondard,1 Rolando Armijo,1 Geoffrey C. P. King,1 Bertrand Meyer2
and Fre´de´ric Flerit3
1LaboratoiredeTectonique,IPGP,CNRS-UMR7578,4,placeJussieu,75252ParisCedex05,France.E-mail:[email protected]
2UPMC,CNRS-UMR7072,Paris,France
3IFG,UniversityofHannover,Germany
Accepted2007August7.Received2007July16;inoriginalform2006February28
SUMMARY
Knowledge on large earthquakes (M ≥ 7.0), geology and fault kinematics is used to anal-
yseconditionsthatfavourisolatedseismicity,clusteredearthquakesorpropagatingsequences
alongtheNorthAnatolianFault(NAF)andtheSeaofMarmarapull-apart.TheoverallNAF–
MarmarafaultsystemisoneofthemostappropriateonEarthtodocumentfaultinteractions
becausereliableinformationcoversalmostcompletelytwoseismiccycles(thepast∼500yr).
Coulombstressanalysisisusedtocharacterizeloadingevolutioninwell-identifiedfaultseg-
ments,includingsecularloadingfrombelowandlateralloadingimposedbytheoccurrenceof
previousearthquakes.EarthquakesalongtheNAFtendtooccurwherepreviouseventshave
increasedthestress,butsignificantisolatedeventsintheSeaofMarmararegion(1894,1912)
have occurred, suggesting the secular loading has been the determining factor. Present-day
loading appears to be particularly high along the 70-km-long segment located in the central
Marmara Sea, southwest of Istanbul. For the 18th century M ≥ 7.0 earthquake clusters, we y
g
construct scenarios consistent with the tectonic and historical data. We find that scenarios o
consistent with slip deficit and secular loading distributions (from below) clearly involve a ol
m
sequencethatpropagateswestwardthroughtheSeaofMarmara,despitethestructuralcom- s
ei
plexity.However,theinferenceofapropagatingsequenceimpliesthateacheventhasoccurred S
inasegmentpreviouslystressedbylateralCoulombstressinteractions.Themostlikelyscenar- I
J
G
iosforthepropagatingsequencearealsoconsistentwithCoulombstressinteractionsbetween
faultswithsignificantnormalslipacrosstheCinarcikbasin.Propagatingearthquakesequences
do not occur every seismic cycle along the NAF. The loading has to be in a particular state
ofstressclosetofailureanduniformallalongthefaultsegmentstoexperiencepropagating
earthquake sequences. Non-uniform stress relief during the 18th century sequence explains
theoccurrenceofisolatedeventsinMarmarain1894and1912.Asaconsequence,thewell-
known 20th century sequence along the NAF has not propagated as a sequence across the
SeaofMarmara.ThemostlinearpartoftheNAFacrossnorthernTurkeybehavesasasingle
faultsegment,accumulatingstressduringhundredsofyearsandrupturingentirelyduringvery
shortperiods.TheMarmarapull-apartfaultsystembehavesasamajorgeometriccomplexity,
stoppingordelayingtheprogressionofearthquakeclusteringandpropagatingsequences.Fault
zonesinteractwitheachotherataverylargescale.
Keywords: earthquakeprediction,Istanbul,Marmara,seismicmodelling,seismotectonics,
stressdistribution.
staticstressinducedbyanearthquake,addedtotheseculartectonic
INTRODUCTION
loading, favours or inhibits the occurrence of subsequent events.
Observationsofearthquakeclusteringandpropagatingsequences Specifically, what are the conditions that favour isolated seismic-
on diverse fault systems have stimulated studies on fault interac- ity, clustered earthquakes or propagating sequences along a fault
tions(e.g.Richter1958;Rice1980;Kasahara1981;Scholz1990). system?AnapproachbasedonCoulombstresschangeshasbeen
Animportantproblemistoassesstowhatextentthevariationof extensivelyusedtoexplainthedistributionofearthquakesinspace
(cid:4)C 2007TheAuthors 1185
Journalcompilation(cid:4)C 2007RAS
1186 N.Pondardetal.
Figure 1. Propagating earthquake sequence along the North Anatolian Fault during the 20th century. (a) Tectonic framework: Extrusion in the eastern
Mediterraneanregion.TheredvectorsrepresenttheGPSvelocitydatareferencedtoafixedEurasia(McCluskyetal.2000).NAF,NorthAnatolianFault;
EAF,EastAnatolianFault;DSF,DeadSeaFault;NAT,NorthAegeanTrough.Thecurvedblackboxisenlargedinb.ThesmallerblackboxlocatestheSeaof
MarmararegionandisenlargedinFig.2.(b)DistributionofearthquakerupturesalongtheNAF(Barka1996;Steinetal.1997;Barkaetal.2002;Akyu¨zetal.
2002;C¸akiretal.2003;Armijoetal.2005).ThemapisanObliqueMercatorprojectionaboutthepoleofrelativemotionbetweentheAnatolianandEurasian
plates,32.6◦E.,30.7◦N.(McCluskyetal.2000).Strike-slipfaultsbetweenthetwoplateslieonsmallcirclesaboutthepoleofrelativemotion;thus,inthis
projectiontheyformhorizontallines(Atwater1970).TheNAFliesonalinerelativelyhorizontal,thedifferentdirectionsofthefaultidentifyingtranstensive
andtranspressiveregions.NNAF,NorthernbranchoftheNorthAnatolianFault;SNAF,SouthernbranchoftheNorthAnatolianFault.(c)Slipdistributionfor
eacheventandthecumulativedisplacement(thickestline)alongthefault.
andtime(e.g.Kingetal.1994;Harris&Simpson1998;Stein1999; ongoingandpastfaultinteractionsalongtheNAF.Acriticalquestion
King&Cocco2000).Themostappropriateplacestoperformanal- istodeterminewhethertheNAFhasexperiencedisolatedevents,
yses of fault interactions are regions where historical and instru- earthquakeclustersorpropagatingsequencesinthepast,especially
mental seismicity exists over several seismic cycles, and where a intheMarmararegion.
well-resolvedfaultgeometryandkinematicsallowsrupturetobe The Sea of Marmara is a major transtensional step-over of the
associatedwithdistinctfaultsegments. NAF(Armijoetal.1999,2002).The1912Ganos(M7.4)and1999
The North Anatolian Fault (NAF) has ruptured during a well- Izmit(M 7.4)earthquakesoccurredattheedgesofthepull-apart
known propagating earthquake sequence between 1939 and 1999 (Fig.2).Theregionhasalsorecordedfivedevastatingearthquakes
(Fig.1)(Tokso¨zetal.1979;Barka1996;Steinetal.1997;Nalbant sincethe18thcentury(1719,1754,1766May,1766August,1894)
et al. 1998). Currently the western tip of the sequence is located (Ambraseys & Finkel 1991, 1995; Ambraseys & Jackson 2000;
intheMarmararegionandaseismicgapremainsclosetothecity Ambraseys 2000, 2001a, 2002). Following the 1999 Izmit event
ofIstanbul.Theobjectiveofthisstudyistodiscussthenatureof studies have been carried out by Hubert-Ferrari et al. (2000) and
(cid:4)C 2007TheAuthors,GJI,171,1185–1197
Journalcompilation(cid:4)C 2007RAS
FaultinteractionsintheSeaofMarmarapull-apart(NorthAnatolianFault) 1187
1912 TMek ir7d.a4g Central Cinarcik Izmit Gulf 12/11/99
Basin Basin Basin 17/08/99
1894 MM~~77
?
Saros Gulf
NNAF
S N A F
Figure2. SegmentationofthefaultsystemandrecentearthquakerupturesintheSeaofMarmarapull-apart(redrawnfromArmijoetal.2002,2005).Thered
circlesindicatethegeometriccomplexities(jogsandfault-bends)thatmaybeassociatedwiththeinitiationandterminationofM≥7.0earthquakeruptures.
Parsonsetal.(2000,2004),buthavebeenhamperedbyalackof thecontinentalcrust(Armijoetal.2003;Hubert-Ferrarietal.2003;
knowledgeofthesubmarinefaultsystemandconsequentuncertain- Fleritetal.2004).
tiesintheearthquakerupturelocations. The NAF is an 800 km long, narrow fault as it crosses north-
Recentlyacquiredbathymetricandseismicdatadefinetheseg- ern Turkey (Fig. 1) (Barka 1992; Hubert-Ferrari et al. 2002), but
mentationoffaultsbeneaththeSeaofMarmara(Armijoetal.2002; becomesmorecomplexintheMarmararegionwhereitsplitsinto
Carton2003;Hirnetal.2003).Theregionalkinematicframework aNorthern(NNAF)andaSouthern(SNAF)branch(Fig.2).The
is constrained using both geological and GPS data (Armijo et al. recentgeologic,morphologic,seismicandGPSresultssuggestthat
1999, 2002; McClusky et al. 2000; Flerit et al. 2003). Detailed slippartitioningcombiningrightlateralmotionandnormalsubsi-
morphologic description of the submarine fault scarps associated denceisresponsibleforthedeformationoftheMarmarapull-apart
with recent major earthquakes has been possible using a Remote region (Armijo et al. 2002, 2005; Carton 2003; Hirn et al. 2003;
OperatedVehicle(ROV).Theseobservationsprovidecriticalinfor- Fleritetal.2003,2004).AtectonicmodelusingGPSvelocitiessug-
mationonthemostrecentearthquakeruptures(1894,1912,1999) geststheAnatolia/EurasiamotionisaccommodatedacrosstheMar-
(Armijoetal.2005;Pondard2006).Hereweidentifypossiblefault mararegionby18–20mmyr–1ofright-lateralslipand8mmyr–1
segmentsthatmayhaverupturedsincethe18thcentury,consistent of extension (Fig. 3) (Flerit et al. 2003, 2004). The normal com-
with macroseismic observations and scaling laws. We model the ponentofmotionismostlydistributedalongtheSNAFwhilethe
evolutionofloadingdistributionsonthosesegmentstodiscusspos- strike-slipcomponentismostlyaccommodatedbytheNNAF.Other
siblescenariosforallM≥7.0earthquakessince1509intheSeaof authorshaveofferedmodelsforthedeformationoftheMarmarare-
Marmararegion.Weassesstherelativeimportanceofthesecular gion (Parke et al. 1999; Le Pichon et al. 2001, 2003) but appear
loadingandthelateralloadingbypreviousearthquakesinCoulomb inconsistentwiththegeologyofthepull-apart(Armijoetal.2002,
interactions. Finally, we compare results of the modelling in the 2005).
MarmararegionwithresultsintherestoftheNAFtodiscussfault
interactionsassociatedwithisolatedevents,earthquakeclustersand
propagatingsequences. SEGMENTATION AND PAST
EARTHQUAKE RUPTURES
The magnitude of the most relevant earthquakes observed in the
Marmararegionisabout7.
TECTONIC SETTING AND
Bathymetric mapping reveals that the main fault strand in the
KINEMATICS: BOUNDARY
SeaofMarmaraisformedbydiscretesegmentssometensofkilo-
CONDITIONS
metres long, separated by geometric fault complexities (jogs and
The formation of the right-lateral NAF results from the process fault-bends)(Fig.2)(Armijoetal.2002).Coseismicslipmeasured
of shear and westward extrusion of the Anatolian plate (Fig. 1a). alongthesesegmentsisabout1–5m(Armijoetal.2005;Pondard
GeologicalstudiessuggestthattheNAFhaspropagatedwestward 2006).Accordingtostrikevariations,eachsegmentaccommodates
throughAnatoliaandtheAegean(Armijoetal.1999).Processes different proportions of strike- and normal-slip. We associate the
ofelasto-plasticfractureassociatedwithlargedamageregionshave initiationandtheterminationofM ≥7.0earthquakeruptureswith
beenproposedtoexplainthelong-termpropagationoffaultsthrough thesegeometricfaultcomplexities,commonlynamedasperitiesand
(cid:4)C 2007TheAuthors,GJI,171,1185–1197
Journalcompilation(cid:4)C 2007RAS
1188 N.Pondardetal.
26˚ 30˚
42˚ Black Sea
u r a s i a
E
Istanbul
20 mm/yr NAF
20 mm/yr Marmara Sea 1122 m mmm/y/yrr 24 mm/yr
F
A
N N Aegean 3 mm/yr
SeaF 2 mm/yr Anatolia
A 4 mm/yr
N
S km
39° 4 mm/yr
5 mm/yr 0 50 100
Figure3. Slip-partitioningintheMarmararegion(Fleritetal.2003).ThevelocityvectorsaremodelledusingtectonicobservationsandGPSdatareferenced
tofixedAnatolia.ThepurplearrowsfollowasmallcirclecentredontheArabia–EurasiaEulerrotationpole.
barriers (e.g. Das & Aki 1977; Kanamori 1978; Aki 1979; King thoughtheycanbeusedinconjunctionwithdetailedstudiesoffault
1983;King&Nabelek1985;Sibson1985). geometrytodefinereasonablerupturescenarios.Onshore,palaeo-
Thelocationandtheslipdistributionofthepastearthquakerup- seismicstudiessuggestthattheIzmitandtheGanosfaultsegments
tures is more or less established. Two categories are distinct: (1) haverupturedduringthe1719andAugust1766events,respectively
theearthquakesforwhichthesurfacebreakhasbeenconstrained (Rockwelletal.2001;Klingeretal.2003).Offshore,themorpho-
withgeologicalandinstrumentalmethods(1999,1912,1894)and logicalstudyofsubmarinescarpssuggestsaslipof4–5massociated
(2) the previous historical events (1719, 1754, 1766 May, 1766 withthe1766Augustearthquake,similartothe1912coseismicslip
August,1894)wheretheextentofrupturemustbededucedfrom (Armijoetal.2005;Pondard2006).Thedistributionofearthquake
thedistributionofhistoricaldamageandaknowledgeoftheactive ruptures associated with the 18th and 19th century events is dis-
faultsystem. cussedinthefollowingsections.
Theslipdistributionassociatedwiththe1999Izmitearthquake
(M 7.4) is well constrained by morphologic and interferometric
COULOMB STRESS MODELLING
observations (Barka 1999; Barka et al. 2002; C¸akir et al. 2003).
The 1912 Ganos earthquake (M 7.4) slip distribution is less cer- Tostudyfaultinteractions,changesofstaticstressassociatedwith
tain.Inbothcasesacriticalproblemhasbeenalackofknowledge largehistoricalearthquakes(M ≥7.0)aremodelledusingrectan-
oftheextensionofruptureintotheSeaofMarmara.Morphologic gulardislocations.Thelengthofruptureandslipforthedislocation
observationsofsubmarinefaultscarpsnowsuggestthatthe1999 sources are chosen to be consistent with seismic moments deter-
earthquake rupture extends 30 km offshore to the eastern tip of minedfrommagnitudeestimates(Ambraseys&Jackson2000;King
the Cinarcik basin and the 1912 earthquake extends 60 km east- etal.2001;Parsons2004)andsegmentlengthsasdiscussedearlier.
wardbetweentheGanosfaultandtheCentralBasinwith4–5mof Thedislocationparametersassociatedwiththemostrecentevents
right-lateralmotion(Fig.2)(Armijoetal.2005).The1894July10 arebasedonmoredetailedfieldandinstrumentalconstraints.The
(M∼7)earthquakemaybecorrelatedwithyoungsubmarinebreaks 1912Ganosearthquakeruptureismodelledusingconstraintsde-
observedattheedgesoftheCinarcikbasin,withsignificantnormal scribedbyArmijoetal.(2005).Theslipdistributionassociatedwith
slip(Armijoetal.2005).Thecorrelationwiththelargebreakalong the1999Izmitruptureisbasedonfieldobservations,morpholog-
theSWedgeoftheCinarcikbasin(∼50kmlong)isconsistentwith icaldescriptionofsubmarinescarpsandSARinterferometricdata
scalinglaws.However,itisnotimpossiblethatthe1894earthquake (Barka1999;C¸akiretal.2003;Armijoetal.2005).Despitepossible
hasrupturedalongtheNEedgeoftheCinarcikbasin,beforethe variationsoflockingdepthindifferentsectorsoftheNAF(around
occurrenceofthesmallfreshbreaksthatareobservedthere,which and across the Sea of Marmara), earthquake ruptures throughout
areprobablyassociatedwiththeMs6.41963event(Armijoetal. the whole region are considered to extend from the surface to an
2005). averagedepthof12km(seediscussioninArmijoetal.2005).A
The location of earthquake ruptures associated with the previ- regionalstressfieldisimposedinagreementwiththekinematicsof
oushistoricaleventsispoorlyknown(1719,1754,1766May,1766 theMarmarapull-apartregiontodetermineoptimumfailuredirec-
August, 1894). However, the Marmara region has been the cen- tions,unlessstrike,dipandrakeofthefaultsarespecified(Nalbant
tre of the Ottoman Empire since 1453 and descriptions of earth- etal.1998;Fleritetal.2003).
quakedamagesincethattimeprovidesubstantialconstraints(seethe ThevariationofstaticstressinducedbyM ∼6earthquakesis
AppendixA)(Ambraseys&Finkel1990,1991,1995;Ambraseys negligiblecomparedtoM∼7eventsforlong-termstressmodelling
2000,2001a,b,2002;Ambraseys&Jackson2000).However,his- (Nalbant et al. 1998; Hubert-Ferrari et al. 2000). For this reason
toricaldataalonecannotdeterminewherefaultruptureoccurred,al- stresschangesaremodelledonlyforearthquakeswithM ≥7.0.
(cid:4)C 2007TheAuthors,GJI,171,1185–1197
Journalcompilation(cid:4)C 2007RAS
FaultinteractionsintheSeaofMarmarapull-apart(NorthAnatolianFault) 1189
for any elapsed time (Figs 4 and 5a). In terms of kinematics, the
earthquakeslipondiscretesegmentsofthefaultrelievespartofthe
slip deficit in some regions and leaves the slip deficit unchanged
in other regions (usually named gaps). A gap may be formed by
one or many segments. Fig. 5 illustrates the distribution of slip
deficits (or loading) in terms of Coulomb stress. For clarity, the
loadingduetosecularloadingthroughoutthefaultandthatdueto
earthquakeslipondiscretesegmentscanbeseparated(Figs5aand
b).Thesuperpositionofthetwoeffects(a+b)isgiveninFigs5(c)
and(d).
Figs 5(a) and (c) show that 4 m of steady slip below 12 km
(correspondingto∼200yrofsecularloadingfortheNAF)creates
critical conditions throughout any segment. Fig. 5(c) shows that
likely Coulomb stress differences due to contrasts in slip deficit
distributionalongthefault(betweenslippedandgapregions)are
large(upto∼80bars).Thisimportantpartoftheloadingappears
simplylinkedtothekinematics.Theothereffectisthatproducedby
rupturedfaultsegmentsinteractingwithneighbouringunruptured
segmentsontheirsides(usuallyreferredasCoulombinteractions).
The corresponding stress induced laterally by a seismic event is
significant(≥40bars)butlimitedtoacertaindistance(≤5km)from
theearthquakerupture’send(Figs5bandc).FromFigs5(c)and(d)
itcanbeappreciatedthatlateralCoulombinteractionsmayplaya
moresignificantrolefortriggeringmoderateeventsrupturingshort
segments(∼10kmlong)thanlongones.Inthefollowingsections
earthquakescenariosandCoulombstressevolutioninMarmaraare
discussedusingthesesimpleconcepts.
SCENARIOS OF 18th AND 19th
CENTURY M ≥ 7.0 EARTHQUAKES
We model possible scenarios for the five M ≥ 7.0 earthquake
Figure4. SeculartectonicloadingmodelalongtheNAFusingtheelastic ruptures in the period from 1509 (after the earthquake) to 1900
dislocationsfromFleritetal.(2003).Theupperpanelshowstheloading (events of 1719, 1754, 1766 May, 1766 August and 1894)
alongallthebranches.Thelowerpanelshowsloadingforthenorthernbranch (Appendix A). No assumption is made for slip deficits result-
alone.ThefaultratesarebasedontheinterseismicGPSvelocities(McClusky
ing from previous events. Although many scenarios can be pro-
etal.2000).Theloadingisimposedstartingtheyear1509.Twohundred
posed, only eight are significantly different and consistent with
yearsofloadingcorrespondstoabout40barsofCoulombstressincrease
tectonic and historical data (Fig. 6). However some loading con-
alongthefaultsegmentsatadepthof5km.Theblackcrossrepresentsthe
ditions due to the instantaneous slip deficit distributions are less
optimumfailuredirections(strike-slipandnormal)(Nalbantetal.1998).
Theprincipalcompressivestressaxisof150–250barsisindicatedbetween probable than others, independently of lateral Coulomb fault
theyellowarrows(orientedN115◦E). interactions.
It is not likely that earthquake rupture is absent along a fault
segmentthatissignificantlyloaded(forinstancemorethan50–80
Ourseculartectonicloadingmodelusesthedislocationmodelof bars).Thisisthecaseforscenarioh(Fig.6)wherenoearthquake
Fleritetal.(2003),whichisbasedoninterseismicGPSvelocities hasoccurredfor500yralongthe60-km-longsegmentatthecentre
andtectonicobservations(McCluskyetal.2000;Armijoetal.2002) oftheSeaofMarmara,suggestingaslipdeficitofmorethan10m.
(Fig.4a).Theelementsforthesecularloadingextendfromalocking Suchaslipdeficitwouldbeenoughtogenerateanimprobablylarge
depthof12kmtogreatdepth.Alikelyrecurrencetimeof200yrof earthquake(M >7.6),unlikelyinviewofthehistoricalseismicity
loadingcorrespondsto40barsofstressincreasealongthelocked (Ambraseys&Jackson2000),thecoseismicslipmeasuredalong
faultsegmentsatadepthof5km.Theloadingisimposedstarting the Sea of Marmara fault segments (1–5 m) (Armijo et al. 2005;
in1509,atthebeginningofaknownquiescentperiodofthenorth Pondard 2006) and earthquake recurrence times deduced for the
Marmararegion(from1509to1719;Ambraseys&Jackson2000). Marmararegion(150–280yr)(Rockwelletal.2001;Klingeretal.
LoadingalongtheSNAFissmallcomparedtothatalongtheNNAF 2003).AllotherpossiblescenariosinFig.6implythatthecluster
and they do not interfere significantly with each other. Moreover of large earthquakes in the 18th century occurred as a westward
changes of static stress due to M ∼ 7 events along the SNAF do propagatingsequencealongthenorthernbranchoftheNAF,across
notinfluencethefaultinteractionsalongtheNNAF(Hubert-Ferrari theSeaofMarmara.
etal.2000).SofaultinteractionswiththeSNAFarenotincluded Conversely,unlessearthquakescanoccurrandomlywithoutany
inourmodels(Fig.4b). relation with loading, a rupture is unlikely to occur along a fault
The kinematic effect of a uniform secular loading (due to slip segmentwheretheslipdeficithasjustbeenrelievedbysignificant
below12km)istheaccumulationofauniformslipdeficitinthe earthquakeslip.Loadingacrosssuchasegmentwouldbetoolowand
lockedregion(above12kmdepth).Thisoccursataratefixedby acertaintimeisrequiredforstresstobuildupagain.Thisappears
thedislocationmodelandaCoulombstressfieldcanbecalculated tobethecaseforscenariosc,d,e,fandg(Fig.6).Anearthquake
(cid:4)C 2007TheAuthors,GJI,171,1185–1197
Journalcompilation(cid:4)C 2007RAS
1190 N.Pondardetal.
Depth (km)
4 m of slip deficit
0
a
5 40 bars from loading
locking10 10 km
depth 4 m of slip
15
+
0
40 bars b
from slip
5 4 m of coseismic slip
10
15
=
4 m of
slip deficit
4 m of slip deficit no slip deficit
0
Future c
event
5 40 bars
80 bars
10
15
0
Future d
5 40 bars event
80 bars
10
15
-100 -50 0 50 100
Coulomb Stress
Change (bars)
Figure5. Loadingdistributionasaresultofsecularloadingandearthquakeslip.(a)Secularloadingassociatedwith4mofsteadyslipbelow12km.Theslip
deficit(4m)isassociatedwith40barsofuniformloadingat5kmdepth.(b)Loadingproducedbyearthquakeslipondiscretefaultsegment(Mw7.1eventwith
4mon35-km-longsegment).Thelateralloadingonadjacentsegmentsislimited(lessthan40barsat5kmdistancefromtherupture).Thesuperpositionof
secularandearthquakeloading(a+b)isgivenin(c)and(d).Forthe35-km-longsegmentinc(widedashedrectangleindicatingfutureeventwithM∼7)the
loadingproducedbysteadyslipbelowcreatescriticalconditionsthroughoutthesegmentwhilethestressduetoearthquakeslipislimited.Forthe10-km-long
segmentind(narrowdashedrectangleforfutureeventwithM ∼6)secularandearthquakeloadingaresimilar.Gridwithdotsandarrowsspecifieszones
wheresliphasoccurredornot.
occurringalongeitherofthefaultswithsignificantnormalslipin theeffectsofrupturepropagationcannotbeincorporated.Although
theCinarcikpull-apartrelievesstressalongtheantitheticsegment. thisisattractiveasaquantitativeapproach,itappearstosufferfrom
Thismaydelaythetriggeringofamajorearthquakeinthebasin. drawbacks.Therawhistoricalinformationhasbeenconvertedinto
Stressinteractionsbetweencloselyspacedantitheticnormalfaults numericalvalues,whichconcealssubstantialuncertainties.Forex-
shouldcauseaneventononefaulttodelayeventsontheother(e.g. ample,aquiescenceofmorethan500yrinthenorthwestSeaof
King&Cocco2000).AsshowninFig.6forscenariose,fandg Marmara deduced from the raw historical data is emphasized by
animprobableearthquakerupturehasfollowedalargeeventinthe Ambraseys&Jackson(2000).Itisnotclearwhythisisnotfound
Cinarcikbasinbyonly12yrandby35yrforscenarioscandd.It bythenumericalprocedureofParsons(2004)usingthesamedata
isunlikelyfortwoM ≥7.0eventstooccurinthebasinwithina set.Observationsofthesubmarinemorphology(Armijoetal.2005)
periodofaboutacentury.Interactionsacrossthebasinmayexplain showclearlythatthemostspectacularscarps(correspondingtothe
alsowhy140yrofloadinghavebeennecessarytotriggeranother 1912andprobablyAugust1766events)areinfactfoundinthenorth-
earthquakeinCinarcik(1894). westSeaofMarmara.Ontheotherhand,thenumericalprocedure
Another scenario has been proposed by Parsons (2004) and is concentratesearthquakesintheeasternSeaofMarmarawherethe
showninFig.7.Theextentofruptureforeacheventisbasedona historicalinformationisrichest.InthescenarioofParsons(2004)
numericalprocedureusingattenuationlawsandhistoricalinforma- fourrupturesoverlapintheCinarcikBasin(1719,1754,May1766
tion.Someeffectsoflocallyincreasedshakingareconsidered,but and1894)justsouthofIstanbul(Fig.7)andnoruptureisidentified
(cid:4)C 2007TheAuthors,GJI,171,1185–1197
Journalcompilation(cid:4)C 2007RAS
FaultinteractionsintheSeaofMarmarapull-apart(NorthAnatolianFault) 1191
Figure6. PossiblescenariosforM≥7.0earthquakerupturesintheSeaofMarmara,between1719and1894.Eachscenarioisconsistentwithreporteddamage
(seeAppendixA)(e.g.Ambraseys&Finkel1991,1995)andobservations(Armijoetal.2005).(a)Scenarioimplyingapropagatingearthquakesequence
rupturingthemainfaultstrandentirely.TheearthquaketimerecurrenceintheCinarcikbasin((cid:2))is140yr.(b)Ascenariosimilartothescenarioa.However
the1754andthe1894earthquakerupturesareinverted.Itisalsopossiblethatthe1894earthquakerupturereactivatedthe1754break.(c),(d)Otherscenarios
implyingapropagatingearthquakesequence.HoweveranearthquakeoccursintheCinarcikbasinonly35yrafteranotherlargeevent.Scenarioswiththe1894
earthquakeoccurringalongthenorthernbranchofthebasinareequivalent.(e),(f),(g)ForthesescenariosanearthquakeruptureoccursintheCinarcikbasin
only12yrafteranotherlargeevent.(h)Theonlyscenarioimplyingacluster.NoearthquakeruptureoccursinthecentreoftheSeaofMarmara.Ifweassociate
theearthquakeruptureswithotherfaultsegmentsitispossibletoproposeotherscenariosimplyingearthquakeclustering,howevernoneareconsistentwith
tectonicandhistoricaldata.
ina30-km-longfaultstretchinthecentralMarmaraSea.According Izmitwithatotallengthof120–180km.(2)The1754September
toourapproachtheseresultsappeartobeunlikely. 2eventbreaksabout30–60kmofnormalfaultatoneoftheedges
Thus only two scenarios appear consistent with loading condi- oftheCinarcikBasin.(3)Thefaultsegmentslocatedbetweenthe
tions,associatedslipdeficitdistributionsandwithCoulombstress CinarcikBasinandtheGulfofSarosbreakduringthe1766May22
interactions across the Cinarcik basin (scenarios a and b; Fig. 6). andthe1766August5events.Itisnotpossibletodistinguishwhich
Scenariosaandbhavefeaturesincommon(Fig.8):(1)The1719 ofthetwoearthquakerupturesextendsfromtheTekirdagBasinto
May25earthquakeruptureextendstothewesterntipoftheGulfof theCentralBasin.Thebreaksassociatedwiththeseeventsareabout
(cid:4)C 2007TheAuthors,GJI,171,1185–1197
Journalcompilation(cid:4)C 2007RAS
1192 N.Pondardetal.
May 1766
1754
1719 1719
August 1766
1894
Figure7. 18thand19thcenturyM >7earthquakerupturesproposedby
Parsons(2004),basedonanumericalprocedureusingattenuationlawsand 1754
historicalinformation.
70and140kmlong,respectively,inscenarioa,or120and50km
longinscenariob(Fig.6).(4)The1894July10earthquakebreaks
30–60kmoffaultsegmentatthenorthernorsouthernedgeofthe
CinarcikBasin,reactivatingthe1754earthquakebreak,orrupturing
theantitheticnormalfault.Fig.9showsdetailsoftheloadingevo-
lutionforscenarioa,andFig.10depictstheresultingpresent-day
May 1766
loading.
COULOMB STRESS EVOLUTION:
FAULT INTERACTIONS VERSUS SLIP
DEFICITS AND SECULAR LOADING
Any possible earthquake scenario consistent with slip deficit and
secularloadingdistributions(Fig.6;scenariosatog)impliesasys-
tematicwestwardpropagationoffaultruptureduringthe18thcen-
turysequence.Therefore,the18thcenturyearthquakesequencehas
veryprobablybrokenthesubmarinefaultsystementirely,inspiteof
August 1766
geometricandkinematiccomplexitiesassociatedwiththeMarmara
pull-apart (Figs 6, 8 and 9). This has important implications for
faultinteractions.Besides,onlytwoscenariosareconsistentwith
CoulombstressinteractionsacrosstheCinarcikbasin(scenariosa
andb).Theyimplythatduringthepropagationonlyoneearthquake
rupture(nottwo)hasoccurredintheCinarcikbasin.
Itisclearhoweverthatthe1894,1912and1999Augustearth-
quakes have not formed any propagating sequence in the Sea of
Marmara (Figs 2 and 11a). In particular, the 1894 and the 1912
events appear as isolated breaks that have not been triggered in
a region stressed by a large nearby event (Figs 9f and g). How-
ever, the 1912 earthquake may be coupled with a smaller Ms 6.9
1894
eventthatoccurredintheGulfofSarosin1893(seeAppendixB)
(Ambraseys & Finkel 1991; Ambraseys & Jackson 2000;
Ambraseys2002).Togetherthesetwoearthquakesmaybeseenasa
singlelargeevent.Thustheoccurrenceofthe1894eventandofthe
possiblycoupled1893–1912eventssuggeststhatstresshasreached
failurecriteriamostlybecauseofseculartectonicloadingalone.So
the1894and1912(plus1893?)eventsmayberegardedasexamples
oflargeisolatedearthquakesalongtheNAF. Figure8. Themostplausiblelocationofearthquakerupturesbetween1719
Thepresent-dayloadingintheMarmararegiondeducedfromour and1894,consistentwithmorphologicobservations(Armijoetal.2005),
preferredscenario(Fig.10a)showsthatCoulombstressishigh(>40 historicaldata(e.g.Ambraseys&Finkel1991,1995)anddistributionofslip
bars,red)intwolargeisolatedregions.OnespreadsovertheGulfof deficitatanyelapsedtime.ThiscorrespondstoscenariosaandbinFig.6,
whichonlydifferinthepreciselocationofthe1754and1894eventsinthe
Saros,theothercoversthemostlystrike-slipsegmentatthecentre
Cinarcikbasin.
oftheSeaofMarmara.AsidefromSarosandcentralMarmara,the
other segments in the Marmara region appear unloaded (white),
while minor loaded zones, positive (red) and negative (blue), are
(cid:4)C 2007TheAuthors,GJI,171,1185–1197
Journalcompilation(cid:4)C 2007RAS
FaultinteractionsintheSeaofMarmarapull-apart(NorthAnatolianFault) 1193
Figure9. DistributionofslipdeficitintheSeaofMarmarabetween1719and1999.ThemodelcorrespondstothescenariopresentedintheFig.6(a).
seenatfaultcomplexitiesbetweensegments.Thisoverallresultis Fig.10(b)incorporatesthecorrespondingCoulombstressrelease
consistentwithearlierCoulombstresscalculationsthatweremade of all earthquakes in the Gulf of Saros. It shows that the central
for the same rate of tectonic loading while superposed with only Marmarasegmentisthemostprobablesiteofruptureforthenext
threerecentlargeevents(1894,1912and1999),thuswithoutcon- largeearthquakeonthenorthernbranchoftheNAF,nexttoIstanbul.
sideringthe18thcenturysequence(Armijoetal.2005).Sothe18th Itwouldoccurasanisolatedevent(notinsequence),howeverasa
centurysequencedoesnotappeartoinfluencemuchtheoccurrence likelyeffectofthelarge-scale20thcenturypropagatingsequence.
oflaterevents,suggestingthatthetectonicloadingwasmostlyre- Itseemsthatpropagatingearthquakesequencesdonotoccurev-
lievedafterruptureofthesequence. eryseismiccyclealongafaultsystem.Specialconditionsofloading
The stress along the fault segment north of the Cinarcik basin arenecessarytotriggerthem.Theloadinghastoberelativelyuni-
appears reduced. This relief is an effect of the 1894 earthquake form and close to failure to trigger earthquake clustering along a
(Ms∼7.3),whichoccurredonacloselyspacednormalfault.Mostof faultzone.Propagatingearthquakesequencesformoneparticular
thepreviousstudiesconcludedtheoppositebecausetheyconsidered categoryofearthquakeclustering.Suchpropagatingsequencemay
thestresstransferassociatedwiththe1999Izmitearthquakeonly occur when the state of loading is so uniform that the variations
(Parsons 2004), or they associated the 1894 earthquake with an ofstaticstressinducedbyanearthquake(∼1–10barsatdistances
onshorefaultsegment(Hubert-Ferrarietal.2000),inconsistentwith ofsometensofkilometresfromtheruptureedge,seeFig.5)are
therecentmorphologicobservations(Armijoetal.2005). sufficienttosuccessivelytriggerevents.
The seismicity in the Gulf of Saros during the 20th and 19th Eventsateachsideofaslip-deficitgaparegenerallyassociated
centuriesincludeseventsofmoderatemagnitudeandforprevious with the loading in the gap. However for long fault stretches the
centurieseventsmuchlessconstrainedbyobservationsthaninthe influenceofeventsonitssidesislessimportantthansecularload-
Sea of Marmara region. However five significant earthquakes are ing produced by the steady slip below the segment (Fig. 5). It is
likelytohaverupturedfaultsegmentsintheGulfofSaros,in1623, because the extent of stress due to a dislocation surface depends
1659,1859,1893(seeAppendixB)and1975(Taymazetal.1991). mainlyonitsshortestdimension.Faultshaveadowndipwidththat
TheMs7.2eventin1659appearsthelargest(Ambraseys2002). isrelatedtothethicknessoftheseismogeniccrust(∼12kminthis
(cid:4)C 2007TheAuthors,GJI,171,1185–1197
Journalcompilation(cid:4)C 2007RAS