ARTICLE IN PRESS · uncertainties with geologic estimates during the most recent phase of the geologic evolution of the E Mediterranean region (post-Late Miocene). On this basis,

Geodetic constraints on the tectonic evolution of the Aegean region and strainaccumulation along the Hellenic subduction zone

Robert Reilinger a,!, Simon McClusky a, Demitris Paradissis b,1, Semih Ergintav c,2, Philippe Vernant d,3a Department of Earth, Atmospheric, and Planetary Sciences, MIT, Cambridge, MA, USAb Higher Geodesy Laboratory, National Technical University, Athens, Greecec Tubitak Marmara Research Center, Earth and Marine Sciences Research Institute, Gebze, Turkeyd Laboratoire Géosciences Montpellier, CC060, Université Montpellier II, Montpellier, France

a b s t r a c ta r t i c l e i n f o

Article history:Received 15 July 2008Received in revised form 29 January 2009Accepted 28 May 2009Available online xxxx

Keywords:Aegean neotectonicsGlobal Positioning SystemGeodynamicsEarthquake hazards

We present evidence that GPS velocity estimates of plate motions and fault slip rates agree to withinuncertainties with geologic estimates during the most recent phase of the geologic evolution of the EMediterranean region (post-Late Miocene). On this basis, we use the GPS differential velocities to estimatethe timing of initiation of the principal structures in NW Turkey, the N Aegean Sea, and central Greece,including, the Marmara Sea, the Gulfs of Evia (GoE) and Corinth (GoC), and the Kephalonia Transform fault(KTF). We interpret these ages to indicate that the North Anatolian fault propagated across the N Aegean,opening the GoE and GoC and initiating the KTF, during the past 1–4 Ma. We further suggest that Aegeanextension that was earlier more distributed across the Aegean Basin became focused on this new fault systemallowing the southern Aegean and Peloponnisos to translate SW with little internal deformation, as observedtoday with GPS. This change in tectonic con!guration may account for the clear geologic evidence for crustalthinning throughout the S Aegean in apparent contradiction with low present-day strain rates. We furthershow that the low present-day strain rate along the southern edge of the Aegean micro-plate requiressubstantial aseismic slip along the plate interface below Crete, consistent with the low level of historic,subduction-type earthquakes along this segment of the subduction zone.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

Understanding the relationship between geodetically determinedsurface motions and longer-term motions estimated from geologicobservations is a prerequisite to using highly precise, present-daygeodetic motion estimates to constrain the Earth's recent geologicevolution. Since the earliest geodetic studies of plate motions, earthscientists have demonstrated good agreement between recent geologic(2–3Ma) and geodetic (~10 yr) platemotion estimates (e.g., Smith et al.,1990; Larson et al.,1997; Sella et al., 2002;McCluskyet al., 2003; Kreemeret al., 2003). However, in themore complex zones of deformationwithinplate interiors and along plate boundaries where geologic estimates ofmotion rates are more dif!cult to determine, the relationship betweengeodetic and geologic motions is less clear.

Geodetic observations in and around the Aegean Sea indicate that atpresent, the southern Aegean Sea and Peloponnisos Peninsula are

translating towards the SW with little internal deformation (McCluskyet al., 2000; Kahle et al., 2000; Reilinger et al., 2006). This present-daymotion appears inconsistent with geophysical and geological evidencefor widespread post-Pliocene extension throughout the region (e.g.,Makris, 1978; Armijo et al., 1996). This apparent inconsistency is at oddswith evidence for agreement, within uncertainties, between GPS andgeologic estimates of plate motions and fault slip rates within the EMediterranean region (McClusky et al., 2000, 2003; Allen et al., 2004;Reilinger et al., 2006).

Here, we present evidence for a tectonic scenario to reconcile thisapparent contradiction. We show that the present tectonic structure oftheMarmara Sea, theGulfs of Evia and Corinth in central Greece, and theKephelonia Transform fault would have developed in the past fewmillion years at present-day, geodetic rates. We suggest that thesestructures result from the propagation of the NAF across the N Aegeanduring the past few Ma (Armijo et al., 1999; McClusky et al., 2000) andthat this tectonic event concentrated Aegean extension in the N Aegeanallowing the S Aegean and Peloponnisos, formerly areas of distributedextension, to translate SW with little internal deformation as observedtodaywithGPS.We further show that the low level of observed present-day strain within the Aegean plate and along the Hellenic subductionzone requires that subduction is occurring in large part aseismically, assuggested independently from the low level of historic, subduction

Tectonophysics xxx (2009) xxx–xxx

! Corresponding author. Fax: +1 617 253 6385.E-mail addresses: [email protected] (R. Reilinger), [email protected]

(D. Paradissis), [email protected] (S. Ergintav), [email protected](P. Vernant).

1 Fax: +30 2107722670.2 Fax: +90 262 6412309.3 Fax: +33 4 67 14 36 42.

TECTO-124631; No of Pages 9

0040-1951/$ – see front matter © 2009 Elsevier B.V. All rights reserved.doi:10.1016/j.tecto.2009.05.027

Contents lists available at ScienceDirect

Tectonophysics

j ourna l homepage: www.e lsev ie r.com/ locate / tecto

ARTICLE IN PRESS

Please cite this article as: Reilinger, R., et al., Geodetic constraints on the tectonic evolution of the Aegean region and strain accumulationalong the Hellenic subduction zone, Tectonophysics (2009), doi:10.1016/j.tecto.2009.05.027

http://dx.doi.org/10.1016/j.tecto.2009.05.027

mailto:[email protected]





http://www.sciencedirect.com/science/journal/00401951

earthquakes along the Hellenic Arc (Jackson andMcKenzie, 1988; Shawet al., 2008).

1.1. Tectonics of the Aegean region

Thepresent-day tectonic frameworkof theAegean region is illustratedschematically in Fig. 1A, and the GPS-derived velocity !eld in Fig. 1B. The

tectonic history of the region is complex involving a long history of north-dipping subduction, an early (Oligocene?) stage of continental collision(mico-continental blocks), followed by recent (Miocene–Pliocene?)extension (see Jackson, 1994 for review). The tectonics are thought tobe in large part a result of subduction-related processes, with continentalcollision being ascribed to the “rafting in” of continental fragments, andextension being related to slab roll-back (Le Pichon and Angelier, 1979;

Fig. 1. A. Schematic tectonic/topographic/bathymetric map of the E Mediterranean. Abbreviations: Gulf of Evia (GoE), Gulf of Corinth (GoC), Kephalonia Transform fault (KTF), NorthAnatolian fault (NAF), East Anatolian fault (EAF), Dead Sea fault (DSF), Marmara Sea (MS), North Aegean Trough (NAT), Peloponnisos Peninsula (Pel). B. GPS-derived velocities withrespect to Eurasia (velocity !eld has been decimated for clarity). GPS velocity uncertainties are 95% con!dence ellipses (modi!ed from Reilinger et al., 2006).

2 R. Reilinger et al. / Tectonophysics xxx (2009) xxx–xxx

ARTICLE IN PRESS



Royden,1993). Aegean tectonics are also thought to be intimately relatedto regional tectonic processes including the collision of Arabia withEurasia, and presumed extrusion of the Anatolian plate (Sengor et al.,1984), although the relationship between these different tectonic events/processes is subject to debate (Gautier et al.,1999).We refer the reader toDixon and Robertson (1984) for amore complete discussion of the broadgeologic evolution of the Aegean region.

1.2. Geodetic and geologic estimates of active deformation in the Aegean

Fig. 2 shows the updated GPS velocity !eld for the Aegean andsurrounding areas in an “Aegean-!xed” reference frame (tabulated inAuxiliary Table 1; velocity!eldmodi!ed fromReilinger et al., 2006). Dataprocessing strategy and error estimation are as described in Reilingeret al. (2006). The Aegean reference frame is determined by minimizing,in a least squares sense, the relative motions between selected GPS siteswithin the Aegean region (selected sites are shown in red in Fig. 2). Thelow level of deformation in the southern and central Aegean and thePeloponnisos is well de!ned. This result was unexpected given the clearevidence for geologically recent deformation of the entire Aegean region(e.g., Jackson, 1994; Armijo et al., 1996; Reilinger et al., 1997).

Well-determined geodetic and geologic estimates are available forArabia platemotion aswell as for slip rates onmajor faults in the easternMediterranean region. Chu and Gordon (1998) report the geologicestimates of the Arabian plate motion based on magnetic anomalies inthe Red Sea covering the past 3 Ma. The 3-Ma Euler vector (Arabia–Nubia) is consistent with the GPS Euler vector in location and rotationrate at the 1 standard deviation level (McClusky et al., 2003).

Fig. 3 shows a comparison of geodetic and geologic estimates offault slip rates in the greater E Mediterranean region (modi!ed fromReilinger et al., 2006). Geodetic slip rates are determined from theelastic block model described by Reilinger et al. (2006). This

Fig. 2. GPS velocity !eld and 95% con!dence ellipses with respect to the central Aegean region (sites shown in red are used to determine the Aegean-!xed reference frame). The lowstrain rate within the central Aegean and Peloponnisos is evident. (For interpretation of the references to color in this !gure legend, the reader is referred to the web version of thisarticle.)

Fig. 3. Geodetic vs. geologic fault slip, and basin extension rates for major structures inthe eastern Mediterranean and Middle East region. Geodetic slip rates are from anelastic block model described by Reilinger et al. (2006). Uncertainties are 1 sigma.Abbreviations as in Fig. 1A, and PSSF = Pambak–Sevan–Sunik Fault, G = Gulf. SeeReilinger et al. (2006) for geologic references (revised estimate of geologic slip rate forthe NAF from Kozaci et al., 2007, and for the MRF from Walpersdorf et al., 2006).Agreement provides the basis for using GPS motion estimates to interpret regionalgeologic evolution.

3R. Reilinger et al. / Tectonophysics xxx (2009) xxx–xxx

ARTICLE IN PRESS



comparison indicates agreement within uncertainties for almost allstructures, including the Gulf of Corinth (Armijo et al., 1996) that weexamine in more detail below. Only the Main Recent fault (MRF) inIran shows what appears to be a signi!cant difference — this may bedue to a real change in slip rate on the fault or insuf!cient geologic

and/or geodetic observations. We conclude, in agreement with someprior studies (e.g., Allen et al., 2004), that relative plate motions andslip rates on major fault systems indicated by GPS in the EMediterranean region are equal within uncertainties (±10%) to thegeologic rates averaged over the past few million years.

Fig. 4. A. Red lines showing our estimates of total right-lateral offset across the Kephelonia Transform fault (55–70 km) and oblique extension across the Marmara basin (i.e., in theorientation of GPS differential velocities across the basin), and our estimates of the ages of these structures (to nearest 0.5 Ma) based on present-day GPS velocities (insets). For theMarmara Sea, we also estimate basin age based on the GPS-derived NAF slip rate (25±2mm/yr, Reilinger et al., 2006), and our estimate of the along strike length of the basin (115 –

150 km) assuming it is a pure, right-lateral, pull-apart basin. B. Red lines showing our estimates of extension across the western (two estimates are shown) and eastern Gulf ofCorinth (10 – 15 km; ~20 km, respectively) and the Gulf of Evia (10–15 km) and the differential GPS velocities across these structures used to estimate the time these basins wereinitiated (insets). C. As B for the North Aegean Trough. Multiple widths are estimated to indicate uncertainty. Tick marks on width estimates indicate alternate interpretations forestimating basin width. D. Summary of the ages of major structures (to nearest 1 Ma) traversing the N Aegean and central Greece. We interpret these results to indicate thatdeformation associated with the NAF traversed the NAegean and central Greece during the past 1–4Ma. (For interpretation of the references to color in this !gure legend, the readeris referred to the web version of this article.)


ARTICLE IN PRESS



Fig. 4 (continued).


ARTICLE IN PRESS



1.3. Propagation of the North Anatolian fault across the North Aegeanand central Greece

Fig. 4A shows our estimate of right-lateral offset across the KTF (55–70 km) and the oblique opening of the Marmara Sea (80–90 km). GPSvelocities for sites on opposite sides of the Kephelonia fault, and outsidethe zone of elastic strain accumulation indicate a current differential rateof 30±1mm/yr (GPS sites KRTS–XRIS, Fig. 2; see also Hollenstein et al.,2008). If the rate ofmotionhasbeen constant, itwould require1.8–2.3Mato develop the total present-day offset across the Kephelonia Fault.

Similarly, the differential GPS velocity oblique to the Marmara Sea is18±2 mm/yr (Fig. 4A). This rate and the estimated offset projected tothis orientation (80–90 km) suggest that the Marmara pull-apart basindeveloped during the past 4–5.6 Ma. Alternately, we estimate the age ofthe Marmara basin from the observed GPS slip rate on the NAF and thealong-fault length of the basin, assuming the basin is a classic pull-aparton the NAF (e.g., Barka and Kadinsky-Cade, 1988; Flerit et al., 2003). Weestimate a total length for the basin of 115–150 km (Fig. 4A; see also LePichon et al., 2003), and use the 25±2 mm/yr NAF GPS slip rate(Reilinger et al., 2006) to deduce a total age for theMarmarabasin of 4.3–6.5 Ma. This age for the Marmara basin, and presumably for the presentcon!guration of the NAF in NW Turkey, compares well with theestimated age for the NAF fromoffsets on thewestern NAF (Armijo et al.,1999) and more regional estimates from offsets along the length of thefault (Westaway, 1994).

Fig. 4B shows our estimates of the extension across the Gulf ofCorinth (GoC) and the Gulf of Evia (GoE) and the differential GPSvelocities across these structures. For the W GoC we estimate a totalextension of 10–15 km based on the present-day width of the rift, and

the differential GPS rate across the fault of 10±1mm/yr (LIDO–KOUN),yielding an age of 0.9–1.7 Ma. This age for the GoC is consistent withinuncertainties with the age estimated from more detailed geologicobservations along this segment of the Gulf (Armijo et al., 1996). For theeastern GoC we estimate a total extension of ~20 km and a present-daydifferential velocity of 10±2mm/yr yielding an age of ~2Ma. Similarly,for the GoEwe estimate a total extension of 10–15 km, and a differentialGPS velocity of 4.5±1 mm/yr (Fig. 4B), yielding an age in the broadrange of 1.8–4.3 Ma.

Total extension across the North Aegean Trough (NAT) (Fig. 4C) ismore dif!cult to estimate. Basin geometry is considerablymore complexthan the other structures we consider. In addition, it seems likely thattheNAegeanwas subject tomultiple episodes of extension since the lateOligocene (e.g., Dixon and Robertson, 1984). In any case, we use thebroad range of offsets across the deepest basins shown in Fig. 4C toestimate a total age for the well-de!ned basins of the NAT of 1–6 Ma.

Uncertainties on our estimated ages are large due to geometriccomplexities of the basin bounding faults (Armijo et al.,1996), aswell asa lack of direct constraints on the time required to establish thepresumed new fault con!guration (i.e., NAF transecting the NAT). Anytime delay required for the system to develop will cause an under-estimate in the timeof initiationof thepresent fault geometry. Extensionoutboard of the main bounding faults will also cause an underestimateof the age of the basin. Conversely, less than 100% extension within thebasin will cause an overestimate. We do not attempt to estimate theseeffects in our analysis — our expectation is that active extensionoutboard of the main basin bounding faults will tend to offset anyoverestimates of extension within the basins, at least to some extent.Furthermore, the general agreement (±10%) between geodetic and

Fig. 5. Schematic illustration of present-day Aegean tectonics showing coherent motion of the central Aegean and Peloponnisos (pink dotted area) with strain concentrated in the NAegean and Gulfs of Corinth and Evia. The gray dotted area shows the extent of the Anatolian Plate, the blue dotted areas show deformation zones in the SE Aegean (trenchwardmotion relative to the Aegeanmicro-plate) and between the north and south branches of the North Anatolian fault system in the NAegean (see Fig. 2). Blue arrows show the sense ofmotion acrossmajor tectonic structures, and red arrows, rates of micro-platemotion relative to Eurasia. (For interpretation of the references to color in this !gure legend, the reader isreferred to the web version of this article.)


ARTICLE IN PRESS



Fig. 6. Perspective view of themodel for the African–Aegean plate interface along the southern Hellenic Arc. The interface extends to the outer trench (black dots) and dips at an averageangle of 20° below Crete (Meier et al., 2004). The 12, 20, and 40 kmdepth contours of the plate interface aremarked. The coupling of the plate interface (phi) can vary gradually between 0(free slipping) and 1 (fully locked). In the case represented here, the fully locked part of the zone of plate interaction lies below Crete (from 20–40 km depth) and is indicated by red (seepro!le in Fig. 7D). The N–S components of the GPSmotion rates and 1-sigma uncertainties in the Aegean are projected onto the pro!le shown for comparison to themodel results in Fig. 7(dashed lines indicate width of the pro!le). (For interpretation of the references to color in this !gure legend, the reader is referred to the web version of this article.)

Fig. 7. N–S Aegean GPS velocity pro!le crossing Crete and theoretical deformation due to locking different segments of the plate interface. A. no-locking, B. locking between 0 and20 km depth (outer blue segment in Fig. 6), C. locking of the interface between 0 and 40 km (outer blue segment and red segment in Fig. 6), D. locking between 20 and 40 km depth(red segment in Fig. 6), showing full (solid grey) and 20% locking (dashed line), E. topography and bathymetry along the pro!le, and F. earthquake hypocenters along the pro!le(Engdahl and Villaseñor, 2002). These results indicate little (b20%), if any, locking on the plate interface, and consequently a reduced probability of large inter-plate earthquakes onthis segment of the subduction zone.


ARTICLE IN PRESS



geologic fault slip rates (Fig. 3) suggests that the time required toestablish thenew fault geometry is short in comparison to fault ages (i.e.,~10% of the total age of the faults or ~105 yrs. in central Greece). We areencouraged by the consistency of the results from our simple analysisand those from more detailed structural studies of the Marmara basinandNAF (e.g.,Westaway,1994; Armijo et al.,1999), and the GoC (Armijoet al., 1996). In addition, we expect that errors in our estimates areroughly equivalent for the different structures we consider, resulting inmore accurate relative than absolute ages that are more relevant for thepresent study.

Overall, the relationships between present-day, GPS motionestimates and basin structures and total fault offsets (Fig. 4D) areconsistent with the deformation having developed sequentially fromeast to west during the past ~1–4 Ma.

1.4. A simple, conceptual model for the recent evolution of the Aegeanregion

McClusky et al. (2000) recognized the apparent discrepancybetween the GPS evidence for coherent motion of the SW Aegeanand Peloponnisos Peninsula and geological and geophysical evidencefor recent extension. They proposed a model where the entire Aegeanunderwent extension during the initial period of African slab rollback;during this period the Aegean crust thinned and subsided. At somelater time, the North Anatolian fault (NAF) developed, apparently inresponse to the extrusion of the Anatolian micro-plate (Sengor et al.,1985), and propagated across N Turkey and the Marmara regiondeveloping theMarmara pull-apart basin (see also Armijo et al., 1999).The NAF continued to propagate westward crossing the NorthernAegean Sea and Central Greece, developing the Gulf of Corinth andKephelonia Transform fault and completely decoupling the southernand central Aegean and Peloponnisos from the Eurasian plate. At thistime the south and central Aegean and Peloponnisos began translatingSW with little internal stretching. This model, schematically illu-strated in Fig. 5, allows initial extension in the Aegean that developedthe current extended crust, and subsequent translation with littleinternal deformation, as observed today by GPS, thereby reconcilingpresent-day and geologic evidence for recent deformation of theAegean region.

1.5. Strain accumulation along the Hellenic Arc

In addition to the implications for the tectonic evolution of theAegean region, the observed lowGPS strain rates in the southernAegeanprovide constraints on the nature of the interaction between the Africanand Aegean plates along this subduction zone plate boundary. Thelocation of the plate interface has been estimated from seismicobservations (Bohnhoff et al., 2001; Meier et al., 2004). There is asubstantial, apparently aseismic, accretionary prism south of Crete asshown in Fig. 6. Themain plate interface has been estimated to lie southof and beneath Crete, dipping north at an angle of about 20° (red sectionin Fig. 6). The elastic lithospheric block rotations and interseismic elasticstrains associated with active subduction are modeled using DEFNODE(McCaffrey, 2002). Fig. 7 shows the resulting horizontal deformationrate along a N–S pro!le (see Fig. 6 for location of the pro!le) for lockingof different sections of the plate interface. Fig. 7A and B show that theGPS measurements are insensitive to the interseismic elastic strainresulting from the locking of the shallower part of the plate interfacesouth of Crete (0–20 km depth). However, it seems unlikely that thepoorly consolidated sediments within the accretionary prism arecapable of maintaining high stresses (Mascle and Chaumillion, 1998).GPSmeasurements in the Aegean aremainly sensitive to the coupling ofthe plate interface located between 20 and 40 km depth that lies belowCrete and corresponds to the area of highest seismic activity. However,the low strain rate along the leading edge of the Aegean plate precludessubstantial (i.e., greater than ~20%) locking of the interface below Crete

(see Fig. 7D). Based on this analysis, and the absence of signi!canthistoric earthquakes that can be associated with the plate interface, weconcur with previous interpretations of historic seismic observations(Jackson and McKenzie, 1988) that plate convergence is occurringprimarily aseismically, possibly due to the introduction of poorlyconsolidated sediments into the trench and/or the effects of slabrollback on the dynamics of the plate interface. This conclusion isconsistent with the low level of historic seismic activity along the arc, inspite of the high rate of convergence and with a recent study indicatingthat the 365 AD Crete earthquake occurred off the plate interface (Shawet al., 2008).

2. Conclusions

Available evidence indicates that present-day, geodetic deforma-tion in the E Mediterranean is generally equal within uncertainties(±10%) to geologic deformation since initiation of the present platecon!guration in the late Miocene/Pliocene. We use this relationshipand young basin structures and fault offsets to estimate the age ofinitiation of the main structures in the NW Turkey, the N Aegean, andcentral Greece. We show that these structures, including the MarmaraSea pull-apart basin, the Gulfs of Corinth and Evia rift structures, andthe right-lateral Kephelonia Fault developed roughly sequentiallyfrom east to west. We suggest that this sequential developmentresulted from the propagation of deformation associated with the NAFacross the N Aegean during the past 1–4 Ma, and that this change intectonic con!guration resulted in the concentration of extensionaldeformation associated with Africa slab retreat in the N Aegeanallowing the S Aegean and Peloponnisos to translate SW with littleinternal deformation. This scenario reconciles the apparent discre-pancy between geodetically determined present-day coherent motionof the Aegean and Peloponnisos with geologic evidence for post LateMiocene, distributed extension.

We further show that the low level of present-day strainwithin theS Aegean precludes substantial strain accumulation along the Aegean–Africa plate interface below Crete. This observation suggests a lowlikelihood of subduction-type earthquakes along this segment of theplate boundary, consistent with historic seismic observations andwith the well-documented 365 AD Crete earthquake having occurredoff the plate interface (Shaw et al., 2008).

Acknowledgments

We thank Clark Burch!el, Leigh Royden, and Demitis Papanikolaoufor illuminating discussions on the tectonic evolution of the Aegeanregion, and two anonymous reviewers for thoughtful comments thatimproved the manuscript. We are grateful to Rob McCaffrey forproviding us his DEFNODE software and for advise on its implementa-tion. This research was supported in part by NSF EAR-0337497 to MIT.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.tecto.2009.05.027.

References

Allen, M., Jackson, J., Walker, R., 2004. Late Cenozoic reorganization of the Arabia–Eurasia collision and the comparison of short term and long term deformationrates. Tectonics 23. doi:10.1029/2003TC001530.

Armijo, R., Meyer, B., King, G., Rigo, A., Papanastassiou, D., 1996. Quaternary evolution ofthe Gulf of Corinth rift and its implications for the Late Cenozoic evolution of theAegean. Geophys. J. R. Astron. 126, 11–53.

Armijo, R., Meyer, B., Hubert, A., Barka, A., 1999. Propagation of the North Anatolian faultinto the north Aegean: timing and kinematics. Geology 27 (3), 267–270.

Barka, A., Kadinsky-Cade, K., 1988. Strike-slip fault geometry in western Turkey and itsin"uence on earthquake activity. Tectonics 7, 663–684.


ARTICLE IN PRESS


http://dx.doi.org/10.1029/2003TC001530



Bohnhoff, M., Makris, J., Stavrakakis, G., Papanikolaou, D., 2001. Crustal investigation ofthe Hellenic subduction zone using wide aperture seismic data. Tectonophysics343, 239–262.

Chu, D., Gordon, R.G., 1998. Current plate motions across the Red Sea. Geophys. J. Int.135, 313–328.

Dixon, J.E., Robertson, A.H.F., 1984. The geological evolution of the Eastern Mediterra-nean. Geol. Soc. Lond. Spec. Publ. 17, 1984.

Engdahl, E.R., Villaseñor, A., 2002.Global Seismicity: 1900–1999. In: Lee,W.H.K., Kanamori,H.,Jennings, P.C., Kisslinger, C. (Eds.), International Handbook of Earthquake andEngineering Seismology, Part A, Chapter 41. Academic Press, pp. 665–690.

Flerit, F., Armijo, R., King, G., Meyer, B., Barka, A., 2003. Slip-partitioning in the Sea ofMarmara pull-apart determined from GPS velocity vectors. Geophys. J. Int. 154, 1–7.

Gautier, P., Brun, J.-P., Moriceau, R., Sokoutis, D., Martinad, J., Jolivet, L., 1999. Timing,kinematics and causes of Aegean extension: a scenario based on a comparisonwithanalogue experiments. Tectonophysics 315, 31–72.

Hollenstein, Ch., Muller, M.D., Geiger, A., Kahle, H., 2008. Crustal motion anddeformation in Greece from a decade of GPS measurements, 1993–2003.Tectonophysics 449, 17–40. doi:10.1016/j.tecto.2007.12.006.

Jackson, J., 1994. Active tectonics of the Aegean region. Annu. Rev. Earth Planet. Sci. 22,239–271.

Jackson, J., McKenzie, D., 1988. The relationship between plate motions and seismictremors, and the rates of active deformation in the Mediterranean and Middle East.Geophys. J. R. Astron. Soc. 93, 45–73.

Kahle, H.-G., Concord, M., Peter, Y., Geiger, A., Reilinger, R., Barka, A., Veis, G., 2000. GPSderived strain rate !eld within the boundary zones of the Eurasian, African, andArabian plates. J. Geophys. Res. 105, 23,353–23,370.

Kozaci, O., Dolan, J.F., Finkel, R.C., Hartleb, R.D., 2007. Late Holocene slip rate for theNorth Anatolian fault, Turkey, from cosmogenic 36Cl geochronology: implicationsfor the constancy of fault loading and strain release rates. Geology 35, 867–870.

Kreemer, C., Holt, W., Haines, A.J., 2003. An integrated global model of present-day platemotions and plate boundary deformation. Geophys. J. Int. 154, 8–34.

Larson, K.M., Freymueller, J.T., Philipsen, S., 1997. Global plate velocities from the GlobalPositioning System. J. Geophys. Res. 102, 9961–9981.

Le Pichon, X., Angelier, J., 1979. The Hellenic arc and trench system: a key to theevolution of the eastern Mediterranean area. Tectonophysics 60, 1–42.

Le Pichon, X., Chamot-Rooke, N., Rangin, C., Sengor, A.M.C., 2003. The North Anatolianfault in the Sea of Marmara. J. Geophys. Res. 108, 2179. doi:10.1029/2002JB001862.

Makris, J., 1978. The crust and upper mantle of the Aegean region from deep seismicsoundings. Tectonophysics 46, 269–284.

Mascle, J., Chaumillion, E., 1998. An overview of Mediterranean Ridge collisionalaccretionary complex as deduced from multi-channel seismic data. Geo-Mar. Lett.18, 81–89.

McCaffrey, R., 2002. Crustal block rotations and plate coupling. In: Stein, S., Freymueller, J.(Eds.), Plate Boundary Zones. Geodynamics Series, vol. 30. AGU, pp. 101–122.

McClusky, S., et al., 2000. GPS constraints on plate kinematics and dynamics in theeastern Mediterranean and Caucasus. J. Geophys. Res. 105, 5695–5719.

McClusky, S., Reilinger, R., Mahmoud, S., Ben Sari, D., Tealeb, A., 2003. GPS constraints onAfrica (Nubia) and Arabia plate motion. Geophys. J. Int. 155, 126–138.

Meier, T., Rische, M., Endrun, B., Va!dis, A., Harjes, H.-P., 2004. Seismicity of the Hellenicsubduction zone in the area of western and central Crete observed by temporarylocal seismic networks. Tectonophysics 383, 149–169.

Reilinger, R., et al., 1997. Global Positioning System measurements of present-daycrustal movements in the Arabia–Africa–Eurasia plate collision zone. J. Geophys.Res. 102, 9983–9999.

Reilinger, R., McClusky, S., Vernant, P., et al., 2006. GPS constraints on continentaldeformation in the Africa–Arabia–Eurasia continental collision zone and implica-tions for the dynamics of plate interactions. J. Geophys. Res. BO5411. doi:10.1029/2005JB004051.

Royden, L., 1993. The tectonic expression of slab pull at continental convergentboundaries. Tectonics 12, 303–325.

Sella, G.F., Dixon, T.H., Mao, A., 2002. REVEL: a model for recent plate velocities fromspace geodesy. J. Geophys. Res. 107 (B4), 2081. doi:10.1029/2000JB000033.

Sengor, A.M.C., Yilmaz, Y., Sungurlu, O., 1984. Tectonics of the MediterraneanCimmerides: nature and evolution of the western termination of Palaeo-Tethys.In: Dixon, Robertson (Eds.), The Geological Evolution of the EasternMediterranean,1984. Blackwell Scienti!c Pub., Palo Alto, CA, pp. 77–112.

Sengor, A.M.C., Gorur, N., Saroglu, F., 1985. Strike-slip faulting and related basinformation in zones of tectonic escape: Turkey as a case study. In: Biddle, K.T.,Christie-Blick, N. (Eds.), Strike-slip Faulting and Basin Formation: Society of Econ.Paleont. Min. Sec. Pub., vol. 37, pp. 227–264.

Shaw, B., et al., 2008. Eastern Mediterranean tectonics and tsunami hazard inferredfrom the AD 365 earthquake. Nat. Geosci. 1, 268–276. doi:10.1038/ngeo151.

Smith, D.E., et al., 1990. Tectonic motion and deformation from satellite laser ranging toLAGEOS. J. Geophys. Res. 95, 22,013–22,041.

Walpersdorf, A., et al., 2006.Difference in theGPS deformationpattern of north and centralZagros (Iran). Geophys. J. Int. 167, 1077–1088. doi:10.1111/j.1365-246X.2006.03147.x.

Westaway, R.,1994. Present-day kinematics of theMiddle East and easternMediterranean.J. Geophys. Res. 99, 12,071–12,090.


ARTICLE IN PRESS


http://dx.doi.org/10.1029/2002JB001862



http://dx.doi.org/10.1029/2005JB004051

http://dx.doi.org/10.1029/2005JB004051

http://dx.doi.org/10.1029/2000JB000033

http://dx.doi.org/10.1038/ngeo151

http://dx.doi.org/10.1111/j.1365-246X.2006.03147.x

Documents

ARTICLE IN PRESS · uncertainties with geologic estimates during the most recent phase of the geologic evolution of the E Mediterranean region (post-Late Miocene). On this basis,