Earth and Planetary Science Letters 433 (2016) 89–98

Contents lists available at ScienceDirect

Earth and Planetary Science Letters

www.elsevier.com/locate/epsl

Central and eastern Anatolian crustal deformation rate and velocity

fields derived from GPS and earthquake data

N.M. Simão a,∗, S.S. Nalbant a, F. Sunbul a, A. Komec Mutlu b

a School of Environmental Sciences, University of Ulster, Coleraine, UKb Bogaziçi University, Kandilli Observatory, Earthquake Research Institute, Istanbul, Turkey

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

Article history:Received 4 July 2015Received in revised form 21 October 2015Accepted 22 October 2015Available online xxxxEditor: P. Shearer

Keywords:East Anatolian FaultNorth Anatolian Faultseismic hazardstrain-rateArabian–Anatolian convergenceseismotectonics of Anatolia

We present a new strain-rate and associated kinematic model for the eastern and central parts of Turkey. In the east, a quasi N-S compressional tectonic regime dominates the deformation field and is partitioned through the two major structural elements of the region, which are the conjugate dextral strike-slip North Anatolian Fault Zone (NAFZ) and the sinistral strike slip East Anatolian Fault Zone (EAFZ). The observed surface deformation is similar to that inferred by anisotropy studies which sampled the region of the mantle closer to the crust (i.e. the lithospheric mantle and the Moho), and is dependent on the presence or absence of a lithospheric mantle, and of the level of coupling between it and the overlaying crust. The areas of the central and eastern parts of Turkey which are deforming at elevated rates are situated above areas with strong gradients in crustal thickness. This seems to indicate that these transition zones, situated between thinner and thicker crusts, promote more deformation at the surface. The regions that reveal elevated strain-rate values are 1) the Elazig–Bingol segment of the EAFZ, 2) the region around the Karlıova triple-junction including the Yedisu segment and the Varto fault, 3) the section of the NAFZ that extends from the Erzincan province up to the NAFZ-Ezinepazarı fault junction, and 4) sections of the Tuz Gölü Fault Zone. Other regions like the Adana basin, a significant part of the Central Anatolian Fault Zone (CAFZ), the Aksaray and the Ankara provinces, are deforming at smaller but still considerable rates and therefore should be considered as areas well capable of producing damaging earthquakes (between M6 and 7). This study also reveals that the central part of Turkey is moving at a faster rate towards the west than the eastern part Turkey, and that the wedge region between the NAFZ and the EAFZ accounts for the majority of the counter clockwise rotation between the eastern and the central parts of Turkey. This change in movement rate and direction could be the cause of the extensional deformation and respective crustal thinning, with the resulting upwelling of warmer upper mantle observed in tomographic studies for the region between the Iskenderun bay and the CAFZ. The partitioning of deformation into an extensional regime could be the cause of the relatively low levels of strain-rate in the south-west part of the EAFZ and the northern part of the Dead Sea Fault Zone. Finally, using this new compilation of GPS data for the central-eastern part of Turkey, we obtained a new Anatolia–Eurasia rotation pole situated at 2.01◦W and 31.94◦N with a rotation rate of 1.053 ± 0.015◦/Ma.

© 2015 Elsevier B.V. All rights reserved.

1. Introduction

An understanding of plate dynamics and the manner in which active faults form and respond to associated strain accumulation are among the most fundamental aspects of active tectonics. This is not only important for understanding long-term behavior of an active fault zone, but also for identifying faults and/or areas of el-evated seismic hazard.

* Corresponding author.E-mail address: nmmsimao@gmail.com (N.M. Simão).

http://dx.doi.org/10.1016/j.epsl.2015.10.0410012-821X/© 2015 Elsevier B.V. All rights reserved.

The deformation rates and kinematics in the western part of Anatolia, the Marmara and the Aegean Seas, have been thoroughly characterized (e.g., Aktug et al., 2009; Le Pichon and Kreemer, 2010; Pérouse et al., 2012; Reilinger et al., 2006) leaving its central and eastern parts kinematics well constrained (Aktug et al., 2013;Alchalbi et al., 2010; Mahmoud et al., 2013; Ozener et al., 2010;Reilinger et al., 2006; Tatar et al., 2012; Yavasoglu et al., 2011)albeit with few integrated deformation rate studies (Özeren and Holt, 2010; Walters et al., 2014). Here we present a new kinematic model of the eastern boundary and the central part of the Anato-lian plate by interpolating published GPS velocities and seismicity data. Through this integrated approach we aim to clarify the tec-

90 N.M. Simão et al. / Earth and Planetary Science Letters 433 (2016) 89–98

Fig. 1. a) Tectonic map of the central and eastern regions of Anatolia. Bold numbered blocks represent the areas of focus in this study. b) Tectonic setting of the north-eastern part of the Mediterranean region showing the northward movement of both the Nubia and Arabia plates towards Anatolia and Eurasia.

tonic regimes and to provide insights into the earthquake hazard for the region.

2. Tectonic setting

Turkey’s complex tectonic setting is predominantly the result of the continental collision of the Arabian and Eurasian plates in the east, and the convergence of the Nubian and Anatolian plates in the south. The Arabian plate is moving in a north-northeast direc-tion in relation to Eurasia at a rate of approximately 15 mm/yr, while the Nubian plate is moving in a northerly direction at a rate of around 10 mm/yr relative to Eurasia (Le Pichon and Kreemer, 2010; Reilinger et al., 2006; Walters et al., 2014). The differen-tial motion between Nubia and Arabia is mostly accommodated by the left-lateral motion of the Dead Sea Fault Zone (DSFZ) (Fig. 1). The northward motion of Arabia results in the continental collision along the Bitlis–Zagros Suture Zone (BZSZ), and Nubia’s northward motion results in the plunge of oceanic lithosphere under the Ana-tolian plate along the Aegean–Cyprean Arc.

Analysis of seismic data from several studies revealed that the crustal thickening in eastern Turkey was less than expected from the magnitude of the uplift of the plateau north of the BZSZ. Pnand Sn velocity anomalies studies lead to the conclusion that such a crustal lid is not present and that the elevation is a result of mantle temperatures typical at least of those of the asthenosphere. This resulted in a hypothesis that envisages a hot mantle plume under the plateau that supports the existing high elevation in the region (Gök et al., 2007; Sengör et al., 2003).

The continuous northward motion of Arabia results in the west-ward extrusion of the Anatolian block between two conjugate ma-jor transform faults, the dextral North Anatolian Fault Zone (NAFZ) and the sinistral Eastern Anatolian Fault Zone (EAFZ) (Fig. 1). The NAFZ is a 1200 km-long fault zone that runs along the northern part of Turkey, roughly parallel to the Black Sea from Karlıova in the east to the Gulf of Saros in the west, connecting the East Anatolian compressional region to the Aegean extensional region. The EAFZ is a 550 km-long fault zone extending between Karlıova (triple junction with the NAFZ) in the north-east and the city of Kahraman Maras in the south-west, where it meets the DSFZ. It is

roughly a north-east trending, sinistral strike-slip fault zone (Fig. 1) which comprises a series of faults arranged in parallel, each with a different amount of slip, sub parallel or oblique to the general trend. The EAFZ and its adjacent section of the NAFZ have been relatively active in the last century (Ambraseys and Jackson, 1998;Nalbant et al., 2002, 2005) and this activity provides additional evidence to data from seismic studies which show that signifi-cant stress builds in sections of the fault zone. In fact, Nalbant et al. (2002) underlined two highly stressed segments along the EAFZ, a segment extending from south of the city of Kahraman Maras to south of the city of Malatya, and the sector between the east of the city of Elazig and the north east of the city of Bingol. The latter ruptured with an Mw 6.1 earthquake in 2011, validating the forecast. The same study also indicated that the other section could accommodate a larger Mw > 7.0 earthquake in the future, highlighting the importance of understanding the long-term strain accumulation and active tectonics within this area. GPS observa-tions over the last two decades, however, reveal that slab pull along the Hellenic trench is now more important than the push from the collision in eastern Turkey (Le Pichon and Kreemer, 2010;Pérouse et al., 2012; Reilinger et al., 2006; Vernant et al., 2014).

3. Data and methodology

Employing the method of Haines and Holt (1993) we derived continuous velocity and strain-rate fields, without the need to pre-viously define the geometry of the rigid blocks, through interpo-lation of published GPS velocities and seismicity data. We paid particular attention to interactions between the conjugate NAFZ and EAFZ, extending this region to the east, and west to the Tuz Gölü Lake.

The existing, large dataset with its broad spatial coverage, permitted us to reliably derive strain-rate fields and their cor-respondent velocity fields between plate motions and, as in our case, between slowly deforming regions (Pérouse et al., 2012;Reilinger et al., 2006). We used GPS velocity vectors from Reilinger et al. (2006), to which we added the velocity vectors pub-lished by Aktug et al. (2013), Alchalbi et al. (2010), Mahmoud et al. (2013), Ozener et al. (2010), Tatar et al. (2012) and

N.M. Simão et al. / Earth and Planetary Science Letters 433 (2016) 89–98 91

Fig. 2. Compilation of GPS data and respective 95% confidence ellipses used in this study of kinematics and deformation within central and eastern Anatolia.

Yavasoglu et al. (2011) (Fig. 2). Following recommendations from Pérouse et al. (2012), for each study we determined the rota-tion that minimized the RMS between those velocity vectors and the velocity vectors from Reilinger et al. (2006) (Table 1 in Sup-plementary Material). The relatively small original RMS and the comparatively small improvements in the RMS imply that there was no systematic bias between the studies. After a close visual comparison between the original and rotated vectors, all the vec-tors of the other studies were rotated to the Reilinger et al. (2006)frame to create a dataset that was as consistent as possible. In this study we only used velocity vectors with one standard deviations (σ Ve, σ Vn) ≤ 2 mm/yr.

Strain-rate fields were determined using a bicubic Bessel in-terpolation on a curvilinear grid (Holt et al., 1995) through least squares minimization of the GPS observations against a modelled horizontal velocity field subject to the constraint of a minimal strain magnitude. A priori, velocity field anisotropic strain-rate vari-ances were assigned to the strain-rate tensor. Each cell in the curvilinear grid was assigned a variance based on the resulting number of earthquakes obtained from the Global CMT catalogue (Ekström et al., 2012), which were filtered to eliminate earth-quakes with Mw < 5.2 to mitigate for the spatial incomplete-ness (where Mc = 5.2 is the computed Magnitude of complete-ness for the Global CMT catalogue for this area), and which were also de-clustered to remove seismic swarms in an attempt to mitigate for temporal incompleteness (Davis and Frohlich, 1991;Frohlich and Davis, 1990). Areas with higher assigned variances have the propensity to strain at higher rates in the fitting of the GPS velocities. Anisotropy is introduced using the Global CMT fo-cal mechanisms (Fig. 3). It is controlled by the principal axis of deformation derived from the summation of the seismic moment tensors (Savage and Simpson, 1997). Velocity fields were deter-mined and interpolated throughout the grid in the same way as the GPS data, but this time minimizing the modelled horizontal velocity field obtained from the moment tensors (Haines and Holt, 1993). Only the bearing, and not the relative direction, of the prin-cipal axes in the strain-rate field are used to define the anisotropy (Fig. 3). Since the Global CMT catalogue is a catalogue that spans more than 40 years, it contains data of variable quality. Due to this the most reliable moment tensors were chosen using the parame-ter criteria of Frohlich and Davis (1999).

A pre-defined crustal structure was established in order to ob-tain the seismological strain-rate fields. We decided to use a seis-mogenic layer thickness of 30 km, as close to 95% of the region’s seismicity occurs within this depth (Fig. 3b), with a crustal shear

modulus of 3.2.1010 N/m2. To obtain the continuous velocity and strain-rate fields, we defined a computational grid with 0.5◦ by 0.5◦ cells (Fig. 3). The cells located over the Eurasian and Nubian tectonic plates were considered rigid and not deformable. All cells, located in the eastern Mediterranean region and Anatolia, were free to deform. The region covered by our model was much larger than our study area, since the modelled grid stretches from the Black Sea to northern Egypt, and from the Aegean Sea to east-ern Arabia. In this way edge effects were avoided. Rigidity was not included in the model in its strictest sense. In its place, the strength of a cell, a factor which controls cell deformation, was introduced through the use of the strain-rate variances (Haines and Holt, 1993; Holt et al., 1995). Strength can be uniform (all cells deform equally) or non-uniform if some cells are allowed to strain at higher rates in the process of fitting observed velocities. The advantage of the heterogeneous model was that it allowed for strain localization and complexity on some of the main tectonic boundaries and/or more seismically active regions, and too add information to regions with sparse GPS coverage. Variance values were also calibrated so that the heterogeneous model would have a total strain-rate value in the same range as the homogeneous one. In this study a large amount of GPS data was compiled and used, providing excellent coverage of the studied area (Figs. 2 and 1 of the Supplementary Material). This results in small differences be-tween the heterogeneous and homogeneous models, keeping the same patterns of regional variations of low and high strain and with differences in the amplitude values for each cell generally smaller than 10% (Figs. 1 and 2 of the Supplementary Material).

The Eastern Mediterranean and Anatolia domains are subjected to interactions between rigid plate blocks and/or diffuse deforma-tion (Reilinger et al., 2006) and therefore GPS measurements close to major faults and/or block boundaries may contain a significant component of transient deformation such as inter-seismic loading and post-seismic relaxation. Nevertheless, previous studies using this method of inverting geodetic data have shown the possibil-ity to distinguish a posteriori whether an area is moving as a rigid block or undergoing diffuse deformation (Holt et al., 1995;Pérouse et al., 2012). However in using this method distinguish-ing between inter-seismic loading and post-seismic relaxation from the resulting continuous velocity field was found to be impossi-ble. Several attempts (e.g., Hearn et al., 2002; Ergintav et al., 2009;Savage, 1990) to relate long term post-seismic relaxation and inter-seismic loading with GPS observable velocities and known regional deformation were inconclusive. A recent study hints that the effect of the 1939 Erzincan earthquake remains locally persistent around

92 N.M. Simão et al. / Earth and Planetary Science Letters 433 (2016) 89–98

Fig. 3. a) Anisotropic variances presented as a measure of the crustal rigidity and anisotropy and used as a priori information in the deforming grid used in the GPS inversion. Variances were obtained from the number of earthquakes in the GCMT catalogue (Ekström et al., 2012). Anisotropy was obtained from the principal axis of deformation derived from the summation of the most reliably obtained seismic moment tensors (Frohlich and Davis, 1999) available on the GCMT catalogue. b) A histogram of the number of earthquakes as a function of depth. c) Magnitude-frequency relationship for the GCMT catalogue for the region. Mc is the Magnitude of Completeness. d) Final number of earthquakes after filtering out magnitudes < Mc = 5.2 and depths ≥ 30 km.

the Karlıova area (Sunbul et al., submitted for publication). The same study also indicates that there is no large-scale post-seismic deformation throughout the region. Hence it is reasonable to hy-pothesize that long-term motion dominate the resulting velocity fields in the study area.

4. Results and discussion

4.1. Strain-rate fields

Turkey’s regional strain-rate field is dominated by the localised NAFZ dextral strike-slip deformation that extends from the eastern Karlıova region to its western termination in the Gulf of Corinth. This accommodates the relative motions and acts as a boundary between the slow moving Eurasian Plate and the faster, west-ward moving, Anatolian Plate. The central–eastern parts of Turkey’s strain-rate field are therefore heavily influenced by the NAFZ kine-matics and its interaction with the EAFZ. In this context we closely examined three sections that seem to play a significant role in the strain partitioning for this area (Fig. 1).

Section 1 expands from Tuz Gölü Lake across to the Ankara Province. Its most important active structure is the Tuz Gölü Fault Zone (TGFZ), a boundary zone between blocks of contrast-ing deformation (Çemen et al., 1999; Koçyigit and Beyhan, 1998). Strain-rates derived for this region (Figs. 4, 5 and 7) confirm that the TGFZ is an active boundary that attempts to compensate for regional differentials in the direction of deformation (Aktug et al., 2013; Çemen et al., 1999; Dirik and Göncüoglu, 1996;Koçyigit and Beyhan, 1998; Koçyigit and Deveci, 2008). Strain-rates reveal changes in tectonic regime, from transtensional at its southernmost part, to transpressional in the middle, towards transtensional in the northern tip near Bala where the 2005 and 2007 earthquakes occurred (Aktug et al., 2013; Tan et al., 2010)(Fig. 5). However, stronger strain-rate values do not occur in the Bala region. Fig. 4 reveals that stronger strain-rate values up to

∼140 nstrain/yr – with a significant proportion of shear strain (∼90 nstrain/yr) (Fig. 7) – are located between the Ankara and Aksaray provinces. The Ankara region also reveals high strain-rate values of ∼100 nstrain/yr (with a shear of ∼70 nstrain/yr). It is in-teresting to note that, as has been already observed by Aktug et al. (2013), the region to the west of Tuz Gölü reveals a change of strain-rate regime from transtensional to compressional (Fig. 5).

Section 2 is bounded in the south by the Iskendurun Bay, to the east by the Maras triple junction and to the north by the Central Anatolia Fracture Zone (CAFZ) (Fig. 1). The GPS derived strain style confirms that present-day tectonics follows the neotec-tonic trends observed in older studies, such that: the Anatolian–Arabian boundary reveals a transpressional regime due to the oblique convergence of the Arabian Plate towards Anatolia along the EAFZ (Gürsoy et al., 2003; Tatar et al., 2004); along the DSFZ the African–Anatolian boundary reveals a transtensional regime in existence since the Pliocene (Karig and Kozlu, 1990) that ex-tends westwards; the transtensional regime also extends north following the CAFZ consistent with the trends observed in ge-ological records for the last ∼4 million years and which were responsible for the formation of the Tuzla Gölü and Sultansa-zlıgı pull-apart basins in the Kayseri Province (Aktug et al., 2013;Koçyigit and Beyhan, 1998) (Fig. 5). The shear strain component of this oblique tectonic regime propagates southwest with rela-tively stronger values along the Aslantas Fault Zone and south of it (Fig. 7). This area is seismically active (Aktar et al., 2000) with strain-rate values up to ∼80 nstrain/yr (Fig. 4) in the areas of Ceyhan–Yumurtalık, in close proximity to the Adana Mw6.2 earth-quake epicentre in 1998 (Aktar et al., 2000), and in the area of Osmaniye. Strain-rate values increase to ∼100 nstrain/yr to the north of Kozan and in the Iskenderun region. The area around Kayseri reveals strain-rates in the order of ∼130 nstrain/yr and in the step-like south-west bent fracture of the CAFZ the strain is ∼90 nstrain/yr. This is in concordance with the findings of Aktug et al. (2013).

N.M. Simão et al. / Earth and Planetary Science Letters 433 (2016) 89–98 93

Fig. 4. a) Resulting Central and Eastern Anatolian regional strain-rate field. The second invariant of horizontal strain represents the “magnitude” of strain and is defined as √(ε2

xx + ε2yy + 2ε2

xy) where εxx , εyy and εxy are the horizontal components of strain-rate tensor. Strain unit is “nstrain/yr” (10−9/yr). Contours in the map represent the crustal thickness (km) obtained by Mutlu and Karabulut (2011). b) Central and eastern Anatolia’s crustal thickness (redrawn from Mutlu and Karabulut, 2011).

Fig. 5. a) Dilatational amplitude and direction component of the regional strain-rate field. Contours in the map represent the crustal thickness (km) obtained by Mutlu and Karabulut (2011). b) Central and eastern Anatolia’s crustal thickness (redrawn from Mutlu and Karabulut, 2011).

Section 3 is the eastern portion of the NAFZ that extends from its junction with the Ezinepazarı Fault, to where it meets the EAFZ at the Karlıova triple-junction. The section is bounded in the west by the Elazig–Bingöl segment and in the east by the Varto Fault (Fig. 1). It contains the Erzincan province, where the great Mw7.9 1939 earthquake rupture started (Cakir et al., 2014). A complex and very seismically active region, it has a long record of earth-quakes, and a seismic gap on the Yedisu segment (Ambraseys and Jackson, 1998; Nalbant et al., 2002, 2005). Fig. 4 shows that this area is the area that is deforming at the highest rate which is in accord with the work of Ozener et al. (2010). The area around Bingöl is known for its multiple, medium/large (M ≥ 6.0) earth-quakes (Nalbant et al., 2005) and it is deforming at high rates (∼320 nstrain/yr) with a strong shear component (170 nstrain/yr) and some extensional deformation bounded on both sides by rela-tively high levels of compression (Figs. 5 and 7). Although this area is within the EAFZ, the direction of shear deformation seems to fol-low the general NW-SE trend of the NAFZ (Fig. 1 in Supplementary Material). This is interesting as the 2003 earthquake ruptured on

a right-lateral structure located to the north of the city of Bingöl, which is conjugate to the EAFZ. The change in direction of de-formation within the Bingöl region might be caused by: 1) the apparent anti-clockwise rotation of the Karlıova segment relative to the mean direction of the EAFZ; 2) the related opening of the Bingöl Basin and its associated shortening which occurred to the north and north-east (Hubert-Ferrari et al., 2009). The relatively seismically calm Yedisu segment, north-west of Karlıova, reveals a slight decrease in strain-rate values (Fig. 4). In comparison, strain-rate values increase dramatically along the Erzincan segment and the subsequent north-western provinces of the NAFZ up to the Ezinepazarı fault junction. The elevated strain-rate patterns for this area concur with the results obtained by different researchers (e.g., Poyraz, 2015; Tatar et al., 2012; Walters et al., 2014). This section of the NAFZ between Erzincan and Niksar is capable of produc-ing large earthquakes like the 1939 (Mw7.9), 1943 (Mw7.7) and 1992 (Mw6.8), (Ambraseys and Jackson, 1998; Cakir et al., 2014;Fuenzalida et al., 1997). The Ezinepazarı-NAFZ junction region, with the second highest strain-rate value (∼270 nstrain/yr), seems

94 N.M. Simão et al. / Earth and Planetary Science Letters 433 (2016) 89–98

Fig. 6. a) Magnitude of the regional strain-rate field. Contours in the map represent the Pn velocity anomalies obtained by Mutlu and Karabulut (2011). b) Pn velocity anomaly and anisotropy (redrawn from Mutlu and Karabulut, 2011).

to play an important role in how earthquakes ruptured in the past. This highly strained region functioned as a barrier to the rupture of the 1939 Mw7.9 earthquake. That particular earthquake began its rupture in the Erzincan section of the NAFZ and was diverted through the Ezinepazarı fault. This was the location of the west-ward ruptures of the 1942 (Mw6.9) and 1943 earthquakes (Cakir et al., 2014). The strong transtensional deformation around the Ezinepazarı-NAFZ junction area might explain why the Erzincan 1939 earthquake ruptured towards the Ezinepazarı fault, this way following the main transpressional regime present in the eastern section of the NAFZ (Fig. 5). The transtensional regime of deforma-tion in the junction area concurs with Tatar et al. (2012) findings.

As mentioned before Nalbant et al. (2002) identified two seg-ments as the likely locations of future damaging earthquakes along the EAFZ based on the Coulomb stress analysis. The first seg-ment extends between the cities of Kahraman Maras and Malatya (Kahramanmaras–Malatya segment), and the second is located on an inferred extension of the EAFZ between Elazig and Bingöl. The 8 March, 2010 Elazig earthquake (Mw = 6.1) ruptured the Elazig–Bingöl segment exactly as was forecasted by Nalbant et al.(2002). This success in forecasting promotes the use of Coulomb stress modelling to identify fault segments which pose a mid-term seismic hazard. It also draws attention to the Kahramanmaras–Malatya segment, which was forecasted as capable of producing Mw > 7.0 earthquake (Nalbant et al., 2002). The present study indicates that these two segments are under different tectonic regimes and accumulating different strain-rates as a result. The Elazig–Bingöl segment is under a quasi N-S compressional regime at its extremities, and an E-W transtensional regime in the middle, with a compressional/tensional strain ratio of ∼2. In comparison the Kahramanmaras–Malatya segment is moving slower than the Elazig–Bingöl segment, at a factor of 4 (Kahramanmaras–Malatya = ∼4 mm/yr, Elazig–Bingöl = ∼16 mm/yr) and under a rela-tively low compressional strain regime (Fig. 5). This means that the Kahramanmaras–Malatya segment has a longer recurrence pe-riod compared to that of the Elazig–Bingöl segment, which could be one of many contributing factors to the relative delay in occur-rence of the forecasted earthquake.

4.2. Comparison with tomographic data

In Section 4.1 the resulting dense GPS array obtained from the compiled data revealed that both Eastern and Central Anatolia ex-

hibit systematic local patterns of internal deformation which are inconsistent with a non-deforming rotating rigid block model. This finding supports the results presented by recent work (Aktug et al., 2013; Özeren and Holt, 2010; Walters et al., 2014). In this section we aimed to explain the internal deformation observed in Central and Eastern Anatolia by contextually comparing it with available tomographic data.

Figs. 4 and 6 compare crustal thickness and mantellic veloc-ity anomalies (lithospheric mantle and asthenospheric) calculated by Mutlu and Karabulut (2011) to effective strain-rates. We also compare crustal thickness to dilatational strain-rates (Fig. 5) and well as anisotropies (Pn) (Mutlu and Karabulut, 2011) to the shear component of the strain tensor (Fig. 7). In collisional tectonics, such as the Arabia–Eurasia collision zone, it is a common prac-tice to assume that the direction of seismic anisotropy, a proxy for the direction of deformation in the upper-mantle, can be asso-ciated with the orientations of lateral shear planes inferred from GPS and tectonic slip rates (Lavé et al., 1996; Özeren, 2012). This assumption requires coupling between the crust and the upper mantle. Furthermore anisotropy has long been used to argue that plate tectonics is driven by mantle convection below the plates, i.e. slab pull and ridge push. In the mantle, anisotropy is normally associated with crystals. Due to their elongate structure, olivine crystals tend to align (lattice-preferred orientation) with the flow and the major strain direction due to mantle convection and/or as a result of tectonic deformation (Kaminski and Ribe, 2002;Ribe, 1992). We gave particular attention to the work of Mutlu and Karabulut (2011) because Pn waves propagate along quasi horizon-tal rays at the Moho and within the lithospheric mantle, and are therefore more likely to effectively sample anisotropy on horizon-tal planes (Baldock and Stern, 2005).

The spatial analysis of Sections 1 and 2 show us that the rela-tive increase in strain-rate values in the Adana Basin, Iskenderun Bay and northern termination of the DSFZ can be related to a strong variation in crustal thickness (Fig. 4). The correlation be-tween the increase in strain-rate values and a strong gradient in crustal thickness variation can also be inferred from the Kayseri section of the CAFZ and southern and central parts of the TGFZ (Fig. 4). The southern part of Section 2 reveals areas of extension at the surface (Fig. 5) through the Iskenderun Bay and extending north up to the step-like fracture of the CAFZ. This extension coin-cides with thin crust (Mutlu and Karabulut, 2011), but more rele-vant is that it also coincides with regions of low velocity anomaly

N.M. Simão et al. / Earth and Planetary Science Letters 433 (2016) 89–98 95

Fig. 7. a) Shear amplitude and lateral direction components of the regional strain-rate field. Contours in the map represent the Pn velocity anomalies obtained by Mutlu and Karabulut (2011). b) Pn velocity anomaly and anisotropy (redrawn from Mutlu and Karabulut, 2011).

in the asthenosphere (down to 200 km depth) (Salaün et al., 2012)and overlaying lithospheric mantle (Fig. 5). Further west between the southern tip of the CAFZ and the TGFZ, the results of Mutlu and Karabulut (2011) reveal that the lithospheric mantle has a very low velocity anomaly and strong levels of anisotropy, with a cor-responding relative increase in values of strain-rate at the surface (Figs. 6 and 7). This indicates that this part of south-central Turkey has a relatively strong crust-lithospheric mantle coupling.

The Karlıova triple junction region (Section 3) and the re-gion to the north/north-east reveal a thicker crust (40–50 km) with a prominent low crustal velocity zone (Gök et al., 2007;Mutlu and Karabulut, 2011). This region lies on top of a por-tion of the upper mantle that shows a relatively high velocity anomaly (Gök et al., 2007; Mutlu and Karabulut, 2011). The re-gions surrounding the junction indicate a lithospheric mantle with low velocity anomalies. The Elazig–Bingöl segment of the EAFZ has very high levels of strain and a change of regime from transpres-sional to transtensional (Fig. 5). It is situated in an area with strong variation in crustal thickness (Figs. 4 and 5), with positive crustal velocity and negative lithospheric mantle velocity anomalies, in contrast to that observed closer to the triple junction by Gök et al.(2007). In the north-west, the Erzincan province also reveals high levels of deformation (Fig. 4). As with the Elazig–Bingöl segment it is a region with a strong gradient in crustal thickness variabil-ity (Fig. 4). The NAFZ-Ezinepazarı fault junction also indicates high levels of strain and a strong variation in crustal thickness. Each of these three highly strained areas are situated in transition zones from thinner to thicker crust, and in areas of strong negative litho-spheric mantle anomalies (Figs. 4 and 6). The region between the Erzincan province and the Ezinepazarı fault junction has strongly negative Pn velocity perturbations (Figs. 6 and 7). This negative anomaly is coupled with pronounced anisotropy, which is simi-lar to the right lateral shear deformation obtained for this area (Fig. 7). Therefore high levels of strain-rate observed at the sur-face along this portion of the NAFZ seem to be connected with the pairing of low velocity perturbation and high levels of anisotropy in the lithospheric mantle. This suggests crust-mantle coupling for the area.

The next step was to compare the directions of both the left-lateral and right-lateral components of the geodetically determined strain-rate for the eastern part of Turkey with Pn anisotropy data obtained from Mutlu and Karabulut (2011) (Fig. 7). We think that the fast-wave polarization of Pn anisotropy in the eastern part of Turkey is mainly dominated by the approximate N-S conver-gence of Arabia towards Eurasia and the resulting westward ex-trusion of Anatolia along the NAFZ and EAFZ. This opinion is based on the observation that lithospheric mantle anisotropy fol-lows a general trend which is similar to the orientations of the NAFZ and the EAFZ, and its corresponding right and left-lateral shear planes inferred from GPS (Fig. 7). This is particularly true at east of ∼36◦E, where anisotropy along the NAFZ and its south-east projection along the BZSZ, follows the right-lateral trend of shear strain computed from GPS. We observed the same for the EAFZ and its north-east projection towards the lower Caucasus re-gion. Anisotropy follows the left-lateral trend of shear strain com-puted from GPS. This common trend between lithospheric mantle anisotropy and the direction of shear deformation at the surface seems to be perturbed specifically in two areas. The first area is the eastern Anatolian plateau between the Lake Van and Karlıova, which was found to lack a lithospheric mantle (Gök et al., 2007). The second area is that which extends from the northern termi-nation of the DSFZ, through Iskendurun Bay, where an upwelling of warm mantle material (Salaün et al., 2012) promotes crustal thinning at the surface. These observations indicate that the N-S convergence of the Arabian plate towards Eurasia and extrusion of Anatolia towards the west, measured at the surface can also be observed along the Moho and within the lithospheric mantle anisotropy. This observation in combination with the finding that low levels of compressional strain (Fig. 5) remain along the EAFZ indicates that the Arabian Plate convergence still plays a role in eastern Anatolia kinematics.

Paul et al. (2014) showed that Anatolia’s underlying upper man-tle shear-wave anisotropy has a counter-clockwise rotation of 1◦per degree of longitude, from eastern Anatolia to the northern Aegean. The authors discuss that this counter-clockwise rotation can be caused by the instantaneous density-driven mantle circula-tion and upwelling from the African super-swell with additional lo-

96 N.M. Simão et al. / Earth and Planetary Science Letters 433 (2016) 89–98

Table 1Anatolia’s Euler vectors relative to Eurasia and 1σ uncertainties.

Reference Lat (◦)

Lon (◦)

�

(◦/Ma)σ(◦)(Lat)a

or (maj)b

σ(◦)(Lon)a

or (min)b

Azimuth (◦) of σ (maj)

σ �

(◦/Ma)

Aktug et al. (2013) 31.68 31.61 1.38 0.05a 0.02a – 0.01Reilinger et al. (2006) 30.8 32.1 1.23 0.8a 0.7a – 0.02Walters et al. (2014) 31.4 32.3 1.14 0.1a 0.1a – 0.01This study 31.94 32.01 1.05 0.14b 0.07b 15.20 0.02

a Uncertainties of Latitude and Longitude.b Major and minor axis of σ ellipse.

Fig. 8. The GPS derived velocity field for Central and Eastern Anatolia. Bold arrows represent the median direction and velocity obtained from the GPS velocity vectors inside each of the bounded rectangles.

cal effects, like slab rollback in the Aegean Sea and a slab window beneath south-western Anatolia. This increase in the rate of rota-tion towards the west is known to strongly influence the extrusion of Anatolia (Pérouse et al., 2012; Vernant et al., 2014) and may be directly responsible, for the extensional behavior and crustal thin-ning observed in Section 2 of Anatolia between the Adana basin and the termination of the DSFZ (Fig. 5).

A special note to the Lake Van and surrounding area, charac-terized by a relatively thick and warm crust (low seismic velocity anomaly) (Gök et al., 2007) above a relatively hot upper mantle (Sengör et al., 2003). This area also shows a strong anisotropy and low velocity perturbation on the lithospheric mantle (Mutlu and Karabulut, 2011) which is not reflected at the surface by strong deformation rates (Figs. 6 and 7). Özeren (2012) suggests that this mismatch is possibly caused by a crust-mantle decoupling along the BZSZ and the region to the north of it.

4.3. Velocity fields

The resulting velocity field (Fig. 8) of eastern Turkey shows the initiation of the anti-clockwise rotation of Anatolia and the in-crease in velocity towards the west of the extrusion of Anatolia from the Arabian–Eurasian convergence (Le Pichon and Kreemer, 2010; Pérouse et al., 2012; Vernant et al., 2014). The East Anato-lian plateau is moving with a median velocity of 14 mm/yr and with a median direction of N16◦W. This slows down to a 7 mm/yr as it approaches the lower Caucasus and changes to a bearing of N4◦E. This change in direction and in velocity of the East Anoto-lian plateau does not increase the strain-rate values in the region. This can be seen as additional evidence that the Arabian–Eurasian convergence is predominantly being accommodated by the extru-sion of Anatolia through the fracture zones. The region of the EAFZ and the region to the south are moving at a median ve-locity of 16 mm/yr with a median direction of N33◦W, while the region to the north, that comprises the eastern part of the NAFZ

and the Ovacik fault, is moving at a median rate of 18 mm/yr with a direction of N58◦W. The Anatolian section in the centre, comprising Iskendurun Bay and the CAFZ is moving at a median rate of 18 mm/yr with direction of N66◦W. This seems to indi-cate that south-eastern Turkey is rotating at a slower rate than the adjacent regions to the north and west, and that the direc-tion of movement changes from N30◦W to nearly N70◦W. This could indicate that the wedge shaped region between the NAFZ and the EAFZ is responsible for most of the rotation. It may cause the extensional regime observed between Iskendurun Bay and the CAFZ. This extension may promote both the crustal and related lithospheric mantle thinning and could be associated with the up-welling of warm material from the upper mantle (Salaün et al., 2012). It may also be causing the slowing of the slip rates in the northern tip of the DSFZ (Alchalbi et al., 2010; Mahmoud et al., 2013). In the north-eastern section of NAFZ (Eurasian side) the ve-locity vectors slow down as the NAFZ approaches the Black Seas (Fig. 8) while the Anatolian side follows the 18 mm/yr, N66◦W pattern described above. This velocity variation, which occurs in a region of strong deformation, coupled with strong tomographic anomalies (Fig. 6), results in the increase in slip rates towards the west observed in this section of the NAFZ (Ozener et al., 2010;Tatar et al., 2012; Yavasoglu et al., 2011). The western part of this study reveals a median movement rate of 20 mm/yr and direction of N77◦W. The effect of this further rotation towards the west and the increase in velocity, can be observed through the increase in strain-rate, and the complex change in deformation regime from transtensional to transpressional, in the TGFZ and the surround-ing area. The resulting velocity field of Eastern and Central Turkey confirms the non-uniform westward movement observed by Aktug et al. (2013) and Reilinger et al. (2006), and as demonstrated in this study, seems to be linked to the upper-mantle dynamics and structure.

Finally, the compilation of GPS data presented in this paper creates the opportunity to obtain a Euler vector of Anatolia in relation to Eurasia using GPS stations that are relatively well dis-tributed all over the Anatolian plate. As a result the AN-EU pole is located at the 32.01◦W and 31.94◦N with a rotation rate of 1.053 ± 0.015◦/Ma (Table 1). The resulting Eulerian vector is com-parable to the results obtained by Reilinger et al. (2006), Aktug et al. (2013) and Walters et al. (2014) within the 1σ confidence level.

5. Conclusions

Geodetically and seismically derived strain-rates have helped us identify which areas of the central and eastern parts of Turkey are deforming at elevated rates (Fig. 4). This allowed us to bring new insights to characterize earthquake hazards for this area, by adding more knowledge to the present-day strain-rates of specific regions. Some of these areas include the Elazig–Bingol segment of the EAFZ; the entire region around the Karlıova triple-junction includ-ing the Yedisu segment and the Varto Fault; the Erzincan province of the NAFZ; and subsequent areas north-west along the NAFZ up to where the NAFZ meets the Ezinepazari fault, and sections of

N.M. Simão et al. / Earth and Planetary Science Letters 433 (2016) 89–98 97

Fig. 9. Geodynamic sketch of Central and Eastern Anatolia. Diagram illustrates that Central and Eastern Anatolian geodynamics are caused by several interacting factors. The influence of Arabia’s convergence on Anatolia’s crustal shear deformation (GPS) and lithospheric mantle anisotropy can be indicative that it is still the main force behind the extrusion in the eastern part of Anatolia. In addition the counter-clockwise rotation observed and linked to the slab-pull along the Hellenic Trench appears to influence the central part of Anatolia. It may be the driving force causing the differential movement rate between the north-eastern and the south-eastern regions and the crustal thinning/extension observed in and around the Iskendurun Bay.

the Tuz Gölü fault zone. Other sections like the Adana Basin, the Ankara province and a significant part of the CAFZ are all deform-ing at smaller but considerable rates, and should be considered as areas capable of producing damaging earthquakes (between M6 and 7).

The provinces of Erzincan and Elazig–Bingol along the NAFZ and the EAFZ respectively are the two most tectonically deforming ar-eas defined by this study. The regions in which they lie have strong gradients in crustal thickness and they border a portion of thick crust lying north-east of the Karlıova triple junction (Fig. 4). An-other example of this is the Iskenderun Bay–Tuz Gölü lake arc that bounds an area of thicker crust and seems to follow the exten-sional deformation seen at the surface. This indicates that these transition zones with thinner crust to thicker crust promote more deformation at the surface.

Özeren (2012) showed us that the low strain-rate levels in the Lake Van area measured by GPS, seem to contradict the strong fast-wave polarization axes of Pn anisotropy in the region, where a Mw > 7.0 earthquake occurred in 2011. The author suggests that this indicates that there is a decoupling of the crust with the mantle within this area. On the other hand, in our study the re-gion north of the Erzincan province, epicentre of the 1939 Mw7.9 earthquake, demonstrates strong negative velocity anomaly and anisotropy along the lithospheric mantle and strong surface defor-mation levels revealing a close coupling between the lithospheric mantle and the crust. This also seems to be true for the region between the southern tip of the CAFZ and the TGFZ where high negative velocity anomalies and strong levels of anisotropy corre-spond with an increase in values of strain-rate at the surface.

In geodynamic terms the eastern part of Turkey is still heav-ily characterized by the Arabia–Eurasia collision (Fig. 9). The quasi N–S compressional tectonic regime creates a deformation field that forms the dominant structural elements of the region, which are the dextral strike-slip NAFZ and the sinistral strike slip EAFZ. This surface deformation is similar, to a degree, to the deformation in-ferred by anisotropy studies which have sampled the mantle region closer to the crust (the lithospheric mantle and the Moho) (Fig. 9). Nevertheless the similarity between geodetic shear deformation and Pn anisotropy is dependent on the degree of coupling between

the crust and the uppermost part of the mantle. Regions with very thin lithospheric mantle, or which completely lack it, as is the case in the eastern Anatolian plateau, do not show this geodetic-seismic anisotropy similarity.

The south-central part of Turkey (Iskenderun Bay–Tuz Gölü Lake) shows differences in amplitude and direction in its counter-clockwise rotation. This can promote the extensional deformation and respective crustal thinning and resulting upwelling of warmer upper mantle observed in tomographic studies. This partitioning of deformation into an extensional regime can be the cause of the relatively low levels of strain-rate in the south-west part of the EAFZ. Finally, using this new compilation of GPS data for the cen-tral, eastern part of Turkey we obtained a new Anatolia–Eurasia rotation pole situated at 2.01◦W and 31.94◦N with a rotation rate of 1.053 ± 0.015◦/Ma.

Acknowledgements

This manuscript profited greatly from discussions and invalu-able help from Nicolas Chamot-Rooke and Gwénaëlle Salaün. We thank the editor Professor Peter Shearer and two anonymous re-viewers, who helped us improve this manuscript. The Generic Mapping Tools was heavily used in the making of the Figures for this paper. Last but not least we would like to thank Marguerite Tarzia for all of her proof reading and very useful suggestions.

Appendix A. Supplementary material

Supplementary material related to this article can be found on-line at http://dx.doi.org/10.1016/j.epsl.2015.10.041.

References

Aktar, M., Ergin, M., Özalaybey, S., Tapirdamaz, C., Yörük, A., Biçmen, F., 2000. A lower-crustal event in the northeastern Mediterranean: the 1998 Adana earth-quake (Mw = 6.2) and its aftershocks. Geophys. Res. Lett. 27 (16), 2361–2364.

Aktug, B., Nocquet, J.M., Cingöz, A., Parsons, B., Erkan, Y., England, P., Lenk, O., Gürdal, M.A., Kilicoglu, A., Akdeniz, H., Tekgül, A., 2009. Deformation of west-ern Turkey from a combination of permanent and campaign GPS data: limits to block-like behavior. J. Geophys. Res., Solid Earth 114 (B10), B10404.

98 N.M. Simão et al. / Earth and Planetary Science Letters 433 (2016) 89–98

Aktug, B., Parmaksız, E., Kurt, M., Lenk, O., Kılıçoglu, A., Gürdal, M.A., Özdemir, S., 2013. Deformation of Central Anatolia: GPS implications. J. Geodyn. 67, 78–96, 00003.

Alchalbi, A., Daoud, M., Gomez, F., McClusky, S., Reilinger, R., Romeyeh, M.A., Al-souod, A., Yassminh, R., Ballani, B., Darawcheh, R., Sbeinati, R., Radwan, Y., Masri, R.A., Bayerly, M., Ghazzi, R.A., Barazangi, M., 2010. Crustal deformation in north-western Arabia from GPS measurements in Syria: slow slip rate along the north-ern Dead Sea Fault. Geophys. J. Int. 180 (1), 125–135, 00030.

Ambraseys, N.N., Jackson, J., 1998. Faulting associated with historical and recent earthquakes in the Eastern Mediterranean region. Geophys. J. Int. 133 (2), 390–406, 00312.

Baldock, G., Stern, T., 2005. Width of mantle deformation across a continental trans-form: evidence from upper mantle (Pn) seismic anisotropy measurements. Ge-ology 33 (9), 741–744.

Cakir, Z., Ergintav, S., Akoglu, A.M., Çakmak, R., Tatar, O., Meghraoui, M., 2014. InSAR velocity field across the North Anatolian Fault (eastern Turkey): implications for the loading and release of interseismic strain accumulation. J. Geophys. Res., Solid Earth 119 (10), 2014JB011360.

Davis, S.D., Frohlich, C., 1991. Single-link cluster analysis of earthquake after-shocks: decay laws and regional variations. J. Geophys. Res., Solid Earth 96 (B4), 6335–6350.

Dirik, K., Göncüoglu, M.C., 1996. Neotectonic characteristics of Central Anatolia. Int. Geol. Rev. 38 (9), 807–817.

Ekström, G., Nettles, M., Dziewonski, A., 2012. The global CMT project 2004–2010: centroid-moment tensors for 13,017 earthquakes. Phys. Earth Planet. In-ter. 200–201, 1–9.

Ergintav, S., McClusky, S., Hearn, E., Reilinger, R., Cakmak, R., Herring, T., Ozener, H., Lenk, O., Tari, E., 2009. Seven years of postseismic deformation following the 1999, M = 7.4 and M = 7.2, Izmit-Düzce, Turkey earthquake sequence. J. Geo-phys. Res., Solid Earth 114 (B7), B07403.

Frohlich, C., Davis, S.D., 1990. Single-link cluster analysis as a method to evaluate spatial and temporal properties of earthquake catalogues. Geophys. J. Int. 100 (1), 19–32.

Frohlich, C., Davis, S.D., 1999. How well constrained are well-constrained T, B, and P axes in moment tensor catalogs? J. Geophys. Res., Solid Earth 104 (B3), 4901–4910.

Fuenzalida, H., Dorbath, L., Cisternas, A., Eyidogan, H., Barka, A., Rivera, L., Haessler, H., Philip, H., Lyberis, N., 1997. Mechanism of the 1992 Erzincan earthquake and its aftershocks, tectonics of the Erzincan Basin and decoupling on the North Anatolian Fault. Geophys. J. Int. 129 (1), 1–28.

Gök, R., Pasyanos, M.E., Zor, E., 2007. Lithospheric structure of the continent–continent collision zone: eastern Turkey. Geophys. J. Int. 169 (3), 1079–1088, 00062.

Gürsoy, H., Tatar, O., Piper, J.D.A., Heimann, A., Mesci, L., 2003. Neotectonic defor-mation linking the East Anatolian and Karatas–Osmaniye intracontinental trans-form fault zones in the Gulf of Iskenderun, southern Turkey, deduced from paleomagnetic study of the Ceyhan–Osmaniye volcanics. Tectonics 22 (6), 1067.

Haines, A.J., Holt, W.E., 1993. A procedure for obtaining the complete horizontal motions within zones of distributed deformation from the inversion of strain rate data. J. Geophys. Res., Solid Earth 98 (B7), 12057–12082.

Hearn, E.H., Hager, B.H., Reilinger, R.E., 2002. Viscoelastic deformation from North Anatolian Fault Zone earthquakes and the eastern Mediterranean GPS velocity field. Geophys. Res. Lett. 29 (11), 44.

Holt, W.E., Li, M., Haines, A.J., 1995. Earthquake strain rates and instantaneous rela-tive motions within central and eastern Asia. Geophys. J. Int. 122 (2), 569–593.

Hubert-Ferrari, A., King, G., Woerd, J.v.d., Villa, I., Altunel, E., Armijo, R., 2009. Long-term evolution of the North Anatolian Fault: new constraints from its eastern termination. Geol. Soc. (Lond.) Spec. Publ. 311 (1), 133–154.

Kaminski, E., Ribe, N.M., 2002. Timescales for the evolution of seismic anisotropy in mantle flow. Geochem. Geophys. Geosyst. 3 (8), 1–17.

Karig, D.E., Kozlu, H., 1990. Late Palaeogene–Neogene evolution of the triple junction region near Maras, South-central Turkey. J. Geol. Soc. 147 (6), 1023–1034.

Koçyigit, A., Beyhan, A., 1998. A new intracontinental transcurrent structure: the Central Anatolian Fault Zone, Turkey. Tectonophysics 284 (3–4), 317–336, 00154.

Koçyigit, A., Deveci, S., 2008. Ankara orogenic phase, its age and transition from Thrusting-dominated Paleotectonic period to the strike-slip Neotectonic period, Ankara (Turkey). Turk. J. Earth Sci. 17 (3), 433–459.

Lavé, J., Avouac, J., Lacassin, R., Tapponnier, P., Montagner, J., 1996. Seismic anisotropy beneath Tibet: evidence for eastward extrusion of the Tibetan litho-sphere? Earth Planet. Sci. Lett. 140 (1–4), 83–96.

Le Pichon, X., Kreemer, C., 2010. The miocene-to-present kinematic evolution of the Eastern Mediterranean and Middle East and its implications for dynamics. Annu. Rev. Earth Planet. Sci. 38 (1), 323–351, 00046.

Mahmoud, Y., Masson, F., Meghraoui, M., Cakir, Z., Alchalbi, A., Yavasoglu, H., Yönlü, O., Daoud, M., Ergintav, S., Inan, S., 2013. Kinematic study at the junction of the East Anatolian fault and the Dead Sea fault from GPS measurements. J. Geo-dyn. 67, 30–39.

Mutlu, A.K., Karabulut, H., 2011. Anisotropic Pn tomography of Turkey and adjacent regions. Geophys. J. Int. 187 (3), 1743–1758, 00015.

Nalbant, S.S., McCloskey, J., Steacy, S., 2005. Lessons on the calculation of static stress loading from the 2003 Bingol, Turkey earthquake. Earth Planet. Sci. Lett. 235 (3–4), 632–640.

Nalbant, S.S., McCloskey, J., Steacy, S., Barka, A.A., 2002. Stress accumulation and increased seismic risk in eastern Turkey. Earth Planet. Sci. Lett. 195 (3–4), 291–298.

Ozener, H., Arpat, E., Ergintav, S., Dogru, A., Cakmak, R., Turgut, B., Dogan, U., 2010. Kinematics of the eastern part of the North Anatolian Fault Zone. J. Geodyn. 49 (3–4), 141–150.

Paul, A., Karabulut, H., Mutlu, A.K., Salaün, G., 2014. A comprehensive and densely sampled map of shear-wave azimuthal anisotropy in the Aegean–Anatolia re-gion. Earth Planet. Sci. Lett. 389, 14–22.

Poyraz, F., 2015. Determining the strain upon the eastern section of the North Ana-tolian fault zone (NAFZ). Arabian J. Geosci. 8 (3), 1787–1799.

Pérouse, E., Chamot-Rooke, N., Rabaute, A., Briole, P., Jouanne, F., Georgiev, I., Dim-itrov, D., 2012. Bridging onshore and offshore present-day kinematics of central and eastern Mediterranean: implications for crustal dynamics and mantle flow. Geochem. Geophys. Geosyst. 13 (9), Q09013, 00000.

Reilinger, R., McClusky, S., Vernant, P., Lawrence, S., Ergintav, S., Cakmak, R., Ozener, H., Kadirov, F., Guliev, I., Stepanyan, R., Nadariya, M., Hahubia, G., Mahmoud, S., Sakr, K., ArRajehi, A., Paradissis, D., Al-Aydrus, A., Prilepin, M., Guseva, T., Evren, E., Dmitrotsa, A., Filikov, S.V., Gomez, F., Al-Ghazzi, R., Karam, G., 2006. GPS constraints on continental deformation in the Africa–Arabia–Eurasia con-tinental collision zone and implications for the dynamics of plate interactions. J. Geophys. Res., Solid Earth 111 (B5), B05411.

Ribe, N.M., 1992. On the relation between seismic anisotropy and finite strain. J. Geophys. Res., Solid Earth 97 (B6), 8737–8747.

Salaün, G., Pedersen, H.A., Paul, A., Farra, V., Karabulut, H., Hatzfeld, D., Papazachos, C., Childs, D.M., Pequegnat, C., SIMBAAD Team, 2012. High-resolution surface wave tomography beneath the Aegean–Anatolia region: constraints on upper-mantle structure. Geophys. J. Int. 190 (1), 406–420, 00011.

Savage, J.C., 1990. Equivalent strike-slip earthquake cycles in half-space and lithosphere–asthenosphere earth models. J. Geophys. Res., Solid Earth 95 (B4), 4873–4879.

Savage, J.C., Simpson, R.W., 1997. Surface strain accumulation and the seismic mo-ment tensor. Bull. Seismol. Soc. Am. 87 (5), 1345–1353.

Sunbul, F., Nalbant, S.S., Simão, N.M., Steacy, S., submitted for publication. Investi-gating viscoelastic post-seismic deformation due to large earthquakes in East Anatolia, Turkey. J. Geodyn..

Tan, O., Tapirdamaz, M.C., Ergintav, S., Inan, S., Iravul, Y., Saatçilar, R., Tüzel, B., Taran-cioglu, A., Karakisa, S., Kartal, R.F., Zünbül, S., Yanik, K., Kaplan, M., Saroglu, F., Koçyigit, A., Altunel, E., Özel, N.M., 2010. Bala (Ankara) earthquakes: im-plications for shallow crustal deformation in Central Anatolian section of the Anatolian Platelet (Turkey). Turk. J. Earth Sci. 19 (4), 449–471.

Tatar, O., Piper, J., Gürsoy, H., Heimann, A., Koçbulut, F., 2004. Neotectonic defor-mation in the transition zone between the Dead Sea transform and the East Anatolian Fault Zone, Southern Turkey: a palaeomagnetic study of the Karasu Rift Volcanism. Tectonophysics 385 (1–4), 17–43.

Tatar, O., Poyraz, F., Gürsoy, H., Cakir, Z., Ergintav, S., Akpınar, Z., Koçbulut, F., Sezen, F., Türk, T., Hastaoglu, K.O., Polat, A., Mesci, B.L., Gürsoy, O., Ayazlı, I.E., Çak-mak, R., Belgen, A., Yavasoglu, H., 2012. Crustal deformation and kinematics of the eastern part of the North Anatolian Fault Zone (Turkey) from GPS measure-ments. Tectonophysics 518–521, 55–62.

Vernant, P., Reilinger, R., McClusky, S., 2014. Geodetic evidence for low cou-pling on the Hellenic subduction plate interface. Earth Planet. Sci. Lett. 385, 122–129.

Walters, R.J., Parsons, B., Wright, T.J., 2014. Constraining crustal velocity fields with InSAR for Eastern Turkey: limits to the block-like behavior of Eastern Anatolia. J. Geophys. Res., Solid Earth 119 (6), 5215–5234.

Yavasoglu, H., Tarı, E., Tüysüz, O., Çakır, Z., Ergintav, S., 2011. Determining and mod-eling tectonic movements along the central part of the North Anatolian Fault (Turkey) using geodetic measurements. In: Special Section: Neogene to Recent Tectonics and Block Rotations in the Aegean and Anatolia (Turkey). J. Geodyn. 51 (5), 339–343.

Çemen, I., Göncüoglu, M.C., Dirik, K., 1999. Structural evolution of the Tuzgölü basin in Central Anatolia, Turkey. J. Geol. 107 (6), 693–706.

Özeren, M.S., 2012. Crust-mantle mechanical coupling in Eastern Mediterranean and Eastern Turkey. Proc. Natl. Acad. Sci. 109 (22), 8429–8433.

Özeren, M.S., Holt, W.E., 2010. The dynamics of the eastern Mediterranean and east-ern Turkey. Geophys. J. Int. 183 (3), 1165–1184.

Sengör, A.M.C., Özeren, S., Genç, T., Zor, E., 2003. East Anatolian high plateau as a mantle-supported, North–South shortened domal structure. Geophys. Res. Lett. 30 (24), 8045, 00217.