Block-like plate movements in eastern Anatolia observed by

Transcription

1 Block-like plate movements in eastern Anatolia observed by InSAR Item Type Article Authors Cavalie, Olivier; Jonsson, Sigurjon Citation Block-like plate movements in eastern Anatolia observed by InSAR 2014, 41 (1):26 Geophysical Research Letters Eprint version Publisher's Version/PDF DOI /2013GL Publisher Wiley-Blackwell Journal Geophysical Research Letters Rights Archived with thanks to Geophysical Research Letters Download date 19/08/ :06:32 Link to Item

2 GEOPHYSICAL RESEARCH LETTERS, VOL. 41, 26 31, doi: /2013gl058170, 2014 Block-like plate movements in eastern Anatolia observed by InSAR Olivier Cavalié 1,2 and Sigurjón Jónsson 1 Received 7 October 2013; revised 28 November 2013; accepted 2 December 2013; published 15 January [1] The question whether continental plates deform internally or move as rigid blocks has been debated for several decades. To further address this question, we use large-scale interferometric synthetic aperture radar (InSAR) data sets to study how eastern Anatolia and its surrounding plates deform. We find that most of the deformation is focused at the North and East Anatolian faults and little intraplate deformation takes place. Anatolia is therefore moving, at least its eastern part, as a uniform block. We estimate the slip velocity and locking depth of the North Anatolian fault at this location to be 20 mm/yr and 14 km, respectively. High deformation gradient found near the East Anatolian fault, on the other hand, suggests that little stress is accumulating along the eastern sections of that fault. Citation: Cavalié, O., and S. Jónsson (2014), Block-like plate movements in eastern Anatolia observed by InSAR, Geophys. Res. Lett., 41, 26 31, doi: /2013gl Introduction [2] Plate motions and plate deformation have been extensively studied since the comprehensive theory of plate tectonics emerged in the late 1960s [McKenzie and Parker, 1967; Le Pichon, 1968]. However, while the theory of plate tectonics has been generally accepted for decades, the question of how plates deform is still a subject of debate [Thatcher, 2009]. This is in part because mechanical properties of the continental lithosphere vary significantly from one region to another and also because the Earth s upper crust is mostly brittle and elastic, but not necessarily representative of the behavior of the entire lithosphere. As a result, there are two opposing views that, on one hand, regard the lithosphere as rigid and mostly undeformable except at its borders [Avouac and Tapponnier, 1993] and that, on the other hand, see the lithosphere as a continuum with faults that are mere surface expressions of distributed deformation at depth [England and Mckenzie, 1982]. [3] The eastern Mediterranean area is a zone of complex tectonics where questions on plate motion dynamics and deformation have often been addressed. The regional tectonics is associated with interactions between three major plates, Eurasia, Africa, and Arabia, as well as the smaller Additional supporting information may be found in the online version of this article. 1 King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. 2 Géoazur, Université de Nice Sophia-Antipolis, CNRS, IRD, Observatoire de la Côte d Azur, Valbonne, France. Corresponding author: O. Cavalié, Géoazur, Université de Nice Sophia- Antipolis, CNRS, IRD, Observatoire de la Côte d Azur, 250 rue Albert Einstein, Valbonne, France. (ocavalie@geoazur.unice.fr) American Geophysical Union. All Rights Reserved /14/ /2013GL Anatolian plate. The collision of the Arabian plate into the Eurasian plate in eastern Turkey, the Caucasus and the Zagros, and the westward movement of the Anatolian plate dominate the deformation in this region [McClusky et al., 2000]. GPS data show that the eastern part of the Anatolian plate moves at a rate of 20 mm/yr with respect to Eurasia. This velocity increases progressively in western Anatolia (from 32 ı E) to about 30 mm/yr near the hellenic trench [Nocquet, 2012]. The forces that drive Anatolia s extrusion have been debated, resulting in several different hypothesis, include the following: (i) The indentation of the Arabian plate into Anatolia and the retreat of the hellenic subduction, respectively, push and pull Anatolia westward, respectively. This model implies that plates are rigid in behavior [Heidbach, 2005]; (ii) The difference in gravitational potential energy (GPE) between the high plateau and thus a thicker crust in eastern Anatolia and thinner crust to the west plays a key role [Ozeren and Holt, 2010]. As a consequence, the Anatolian plate moves westward according to the GPE gradient; (iii) Finally, recent studies argue for the existence of a toroidal flow in the asthenosphere that may drag the lithosphere above [Faccenna and Becker, 2010; Le Pichon and Kremer, 2010], and therefore be responsible for the northward motion of the Arabian plate and Anatolia s extrusion. [4] A key point to discriminate between different hypotheses of plate dynamics is to constrain the rheology of the lithosphere. For example, the indentation of the Arabian plate would push Anatolia to the west only if the lithosphere were rigid enough. To improve our understanding of the driving mechanisms of the westward motion of Anatolia, we focus here on the surface deformation near the eastern end of the Anatolian plate where it forms a triple junction with the Eurasian and Arabian plates (Figure 1). This particular area is key to our understanding of how the plates interact across their borders [Wright et al., 2004]. [5] GPS measurements have been the main source of data used to constrain continental lithospheric models in the last 20 years [Thatcher, 2009] and have been very effective in constraining large-scale plate motions and plate boundary deformation. However, in most parts of the world, GPS station density is not sufficiently high to provide information about possible small-scale deformation within a plate or near-plate boundaries. This is indeed the case near the East Anatolian Fault (EAF) [Reilinger et al., 2006] and around the triple junction between Anatolia, Eurasia, and Arabia, although constraining the deformation in this area may provide crucial information for understanding the dynamics of the westward extrusion of the Anatolian plate. We therefore focus on using interferometric synthetic aperture radar (InSAR) data to improve the spatial resolution of the surface deformation in the region. This task poses a significant challenge, because it requires InSAR data processing at the continental scale. 26

3 Figure 1. Tectonic setting and GPS velocities [Nocquet, 2012] in eastern Anatolia. The westward extrusion of the Anatolian plate is bounded by the North and East Anatolian faults (NAF and EAF). (a) The red rectangle shows (b) the area covered. Major fault segments are shown in black, GPS velocities in purple [Ozener et al., 2010], red [Tatar et al., 2012], and blue [Nocquet, 2012], and the coverage of the InSAR data used in this study by the white rectangles. 2. Data and InSAR Method [6] We used multiframe Envisat synthetic aperture radar (SAR) images provided by the European Space Agency to map the surface deformation in eastern Anatolia along both descending and ascending orbits. We first selected tracks with more than 15 images acquired between March and October, but excluded winter scenes to avoid coherence loss due to snow. The selected data are from two descending tracks and one ascending track (Figures 1 and S1 in the supporting information), and from these data, we generated interferograms using the New-Small BAseline Subset (NSBAS) processing chain [Doin et al., 2012]. This chain, based on the ROI_PAC (Repeat Orbit Interferometry Package) software [Rosen et al., 2004], allows, notably, interferograms to be corrected for the stratified tropospheric delay in the data using ERA Interim, a global atmospheric model [Uppala et al., 2005]. In eastern Anatolia, where elevation varies drastically between the plains (300 m in Syria) and the highest mountains (3000 in central Turkey), this correction is a key step to obtaining a successful measurement of tectonic signals by InSAR. Furthermore, we used Doppler orbitography and radiopositioning integrated by satellite orbital and Shuttle Radar Topography Mission topographic [Farr and Kobrick, 2000] information in the processing and then pixel multilooking and adaptive filtering to improve the coherence of the interferograms before unwrapping them using the SNAPHU (Statistical- Cost, Network-Flow Algorithm for Phase Unwrapping) software [Chen and Freymueller, 2002]. Before computing the surface displacement time series, we corrected the interferograms for ramps caused by a drift in the local oscillator onboard the Envisat satellite [Marinkovic and Larsen, 2013]. This drift only causes a ramp in the across-track direction (range) of the interferograms and we therefore made no ramp corrections in the along-track direction (azimuth), to avoid removing tectonic signals related to the motion of the Anatolian and Arabian plates, we estimate the residual orbital signal by an east-west ramp across each interferogram only on their Eurasian part that is considered as stable. We selected stable Eurasia as a reference region and set the mean displacement of this area to zero. [7] To derive surface displacement rates from the interferograms we used small baseline time series analysis. This method is based on using interferograms with small spatial baselines to maximize coherence and the number of pixels to use in the analysis. Our time series analysis is based on constrained least squares inversion to estimate the surface displacement history of each coherent pixel from the set of small baseline interferograms [Doin et al., 2012]. A smoothening operator is applied to limit phase variations due to turbulent atmospheric delays. Finally, we computed the linear component of the time series result for each pixel, as here we are primarily interested in the steady surface displacement velocities. 3. InSAR Results and Modeling [8] The velocity maps estimated from the InSAR data clearly show the motion of the Anatolian and Arabian plates with respect to the Eurasian plate (Figure 2). The motion of Anatolia is particularly strong and appears to be quite homogenous away from the descending (Figure 2a) and toward the ascending look directions (Figure 2b), respectively, reflecting the mostly westward motion of Anatolia with respect to Eurasia. The motion of the Arabian plate is also clear in the descending velocity map but it appears to be mostly stable with respect to Eurasia in the ascending map, as the expected northwest motion of Arabia is perpendicular to the ascending look direction. Limited intra-anatolian deformation is seen in the velocity maps (Figure 2), and most of the short wavelength signals are residual atmospheric delays due to three-dimensional heterogeneities in the atmosphere [Hanssen, 2001]. [9] A velocity transition across the North and East Anatolian faults is clear in both velocity maps. In case of the North Anatolian fault, the velocity transition appears to be fairly gradual with the plate boundary deformation extending several tens of kilometers into the plates (Figure 3d). The transition is much sharper across the East Anatolian fault where a narrow velocity discontinuity is seen between the two plates, at least along the central portion of the imaged fault (Figure 3d). [10] The observed concentration of deformation near major plate boundaries and the lack of intraplate deformation 27

4 Figure 2. Estimated horizontal velocities from InSAR time series analysis in mm/yr parallel to the satellite line of sight (LOS) direction (black and white arrows show the LOS and flight directions, respectively) and in an Eurasia-fixed reference frame. (a) For descending track 264 and 493, velocities are shown positive in the east-southeast direction (103 ı N) and the mean velocities have been computed for the overlapping area. (b) For ascending track 400, velocities are shown positive in the east-northeast direction (77 ı N). support models that describe tectonic plates as elastic blocks rather than a continuum. To simulate the observed velocity field, we therefore adopted the so-called back slip approach, where boundaries between elastic blocks are locked from the surface down to a certain depth (the locking depth), but are allowed to move freely below that depth [Savage and Prescott, 1978]. The eastern Anatolian region is modeled as three blocks separated by two vertical faults, the NAF and the EAF. Several linear segments are used to represent both the NAF and the EAF to account for strike variations along the faults (Figures 3b, S2, and Table 1) and both opening and closing are allowed across each segment, in addition to strike-slip motion. With Eurasia held fixed, we use the InSAR and GPS velocities to estimate the north and west components of the motion of the Anatolian and Arabian plates, as well as the locking depth of the NAF and EAF. We thus search for the values of six model parameters that minimize the misfit between the simulated velocities and the observed InSAR velocities obtained along the three tracks, and the published GPS velocities from this region [Ozener et al., 2010; Tatar et al., 2012]. [11] The low root-mean-square misfit value of 1.6 mm/yr shows that the simulated velocity field matches the observed InSAR and GPS velocities well (Figures 3c and S3c). The velocity of the Anatolian plate, with respect to Eurasia, is estimated to be about 20 mm/yr in a direction parallel to the NAF. The Arabian plate, on the other hand, is found moving to the north-northwest at a rate of 18 mm/yr. This indicates that the motion of Anatolia, with respect to Arabia, is 13 mm/yr. This rate is somewhat higher than rates previously estimated using GPS data from the region, which are around 10 mm/yr [McClusky et al., 2000; Reilinger et al., 2006], but closer to predictions of recent global plate motion models [Demets et al., 2010]. Interestingly, the estimated fault normal velocities for both the NAF and the EAF are low (Table 1), suggesting limited compression or extension across the faults. The east-west profile (Figure 3e) shows that the Anatolian plate velocity decreases slightly near the triple junction because of the locked portion of the faults near the surface, implying that the triple junction moves also westward in time. [12] Estimating the fault-locking depth as well as the interseismic loading rate is essential for assessing the seismic hazard of the area. Here we inverted for a uniform locking depth along each fault to obtain an average value for the thickness of the seismogenic layer in the two cases. We found an optimal locking depth for the NAF of about 14 km, which is similar to the locking depth values that have been previously reported for this fault [Wright et al., 2001; Walters et al., 2011]. The estimated locking depth of the EAF, on the other hand, is much smaller, only 4.5 km, suggesting that either this part of the fault creeps almost freely to the surface or that a narrow compliant zone around the fault accommodates most of the near-surface deformation. Our estimation appears to be robust and not driven by any particular data set, as similar shallow locking depth 28

5 Figure 3. Modeling of the observed velocity field. (a) The horizontal InSAR rate map (from Figure 2a) with GPS velocities shown as black arrows and the NAF and EAF as black lines. Blue and white arrows show the LOS and flight directions, respectively. (b) Elastic back slip model prediction for the InSAR and GPS measurements (red lines indicate the fault model segments) and (c) the residuals between observed and modeled velocities. (d and e) The observed and modeled horizontal velocities along N-S and W-E profiles are shown (profile locations in Figure 3c). The InSAR data (gray dots) are from within 5 km of the profile line. Black lines show median-filtered values and the model prediction is shown in red. Blue dots indicate the profile-projected GPS velocities near the profiles and the associated error bars. values are obtained when using any of the three different InSAR data sets alone in the estimation. A deeper locking depth is found along the EAF further to the west (15 km) [Walters, 2013], indicating that there are significant locking depth variations along the fault, as have been identified along some other faults [Smith-Konter et al., 2011], including along the NAF [Kaneko et al., 2013]. 4. Discussion and Conclusions [13] The small locking depth found along the EAF is surprising because large historical earthquakes (M s > 7.3) have occurred along this segment of the fault [Ambraseys and Melville, 1995; Ambraseys and Jackson, 1998]. After compiling 70 years of seismic data, however, Jackson and McKenzie [1988] pointed out that the seismicity recorded during the twentieth century was low and probably not representative of a longer period. Similarly, Burton et al. [1984] stated, based on analysis of the seismicity between 1900 and 1978, that the seismic activity of the EAF seemed surprisingly low and that it might be an area where seismic creep is dominant. In addition, Bulut et al. [2012] recently found low seismicity rates along the eastern half of the EAF (38 ı E 41 ı E) for the time period Periods of low seismic activity in this area have also been identified in the past 500 years, e.g., during the 250 years from 1544 to 1789 Table 1. Back Slip Model Parameters for Different Segments of the NAF and EAF (Segment Locations are Shown in Figure S2) NAF EAF Segment 1 Segment 2 Segment 3 Segment 1 Segment 2 Segment 3 Length (km) Locking depth (km) Dip Vertical Vertical Strike ( ı N) Strike slip (mm/yr) Extension(+)/compression( ) (mm/yr)

6 [Ambraseys, 1989]. During the following 116 years, a series of relatively large earthquakes (estimated magnitude, M s, between 6.6 and 7.2 [Ambraseys, 1989]) occurred, followed by another period of quiescence. [14] Reconciling the occurrence of both fault creep and large earthquakes is not straightforward. The InSAR observations do not show a sharp displacement discontinuity across the EAF, however, indicating that the fault creep does not reach the surface. Part of the fault may therefore be able to accumulate elastic strain. Also, the shallow locking depth found along the EAF is an average value for a 300 km long segment of the fault. The possibility that particular sections of the fault may accumulate elastic strain cannot be ruled out. And finally, it is possible that the creep is transitory and only active during some periods of time, as suggested by paleoseismological studies [Ambraseys, 1989]. [15] The fault weakness that allows to focus most of the deformation at the Anatolian plate s borders seems to be an essential element in the westward extrusion of the Anatolian plate. Finite element models of the region represent the Anatolian plate boundaries as a weak narrow zones, with lowered viscosity by several orders of magnitude [Faccenna and Becker, 2010], or more realistically [Faccenna and Becker, 2010], as discontinuities with low friction coefficients of around 0.2, to explain the observed westward motion of Anatolia near its eastern end. [16] In conclusion, our results support a tectonic model where plates, in the eastern Anatolian region, behave mostly as rigid blocks and appear to deform primarily near the East and North Anatolian faults. The NAF and EAF are, therefore, weak zones that facilitate the westward extrusion of the entire Anatolian plate in this area. This conclusion applies at the scale of the study (i.e., few hundreds of kilometers), but does not hold at the scale of the entire Anatolian plate (1000 km) as extension clearly occurs in the west. A direct consequence of the block-like deformation is migration of the triple junction to the west as the Anatolian plate velocity only slightly decreases near its eastern end. The InSAR results provide new information about the Arabia- Anatolia plate boundary, including a shallow locking depth with likely creep in the upper crust and a relative motion at the boundary that is faster than previously determined with GPS data. This new short-term (geodetic) rate is in a better agreement with longer-term (geological) rates predicted by global plate motion models like MORVEL (Mid-Ocean Ridge VELocity) [Demets et al., 2010]. [17] Acknowledgments. The SAR data were provided by the European Space Agency through category-1 project We thank Marco Bohnhoff and an anonymous reviewer for their helpful comments. [18] The Editor thanks Gilles Peltzer and Marco Bohnhoff for their assistance in evaluating this paper. References Ambraseys, N. N. (1989), Temporary seismic quiescence: SE Turkey, Geophys. J., 96, Ambraseys, N. N., and J. A. Jackson (1998), Faulting associated with historical and recent earthquakes in the Eastern Mediterranean region, Geophys. J. Int., 133, Ambraseys, N. N., and C. P. Melville (1995), Historical evidence of faulting in Eastern Anatolia and Northern Syria, Ann. Geophys., XXXVIII, Avouac, J.-P., and P. Tapponnier (1993), Kinematic model of active deformation in central Asia, Geophys. Res. Lett., 20(10), Bulut, F., M. Bohnhoff, T. Eken, C. Janssen, T. Kılıç, and G. Dresen (2012), The East Anatolian Fault Zone: Seismotectonic setting and spatiotemporal characteristics of seismicity based on precise earthquake locations, J. Geophys. Res., 117, B07304, doi: /2011jb Burton, P. W., R. Mcgonigle, K. C. Makropoulos, and S. B. Ücer (1984), Seismic risk in Turkey, the Aegean and the eastern Mediterranean: The occurence of large magnitude earthquakes, Geophys. J. R. Astron. Soc., 78, Chen, Q., and J. T. Freymueller (2002), Geodeteic evidence for a near-fault compliant zone along the San Andreas fault in the San Francisco bay area, Bull. Seismol. Soc. Am., 92, Demets, C. R., R. Gordon, and D. Argus (2010), Geologically current plate motions, Geophys. J. Int., 181, 1 80, doi: /j x x. Doin, M.-P., F. Lodge, S. Guillaso, R. Jolivet, C. Lasserre, G. Ducret, R. Grandin, E. Pathier, and V. Pinel (2012), Presentation of the small baseline NSBAS processing chain on a case example: The Etna deformation monitoring from 2003 to 2010 using Envisat data, Proceedings of FRINGE 2011 Workshop, Frascati, Italy, September 2011 (ESA, SP-697, January 2012). England, P. C., and D. Mckenzie (1982), A thin viscous sheet model for continental deformation, Geophys. J. R. Astron. Soc., 70, Faccenna, C., and T. W. Becker (2010), Shaping mobile belts by small-scale convection, Nature, 465, , doi: /nature Farr, T. G., and M. Kobrick (2000), Shuttle radar topography mission produces a wealth of data, Eos Trans. AGU, 81(48), Hanssen, R. F. (2001), Radar Interferometry, Data Interpretation and Error Analysis, 308 pp., Kluwer Academic Publishers, Dordrecht, Netherlands. Heidbach, O. (2005), Velocity Field of the Aegean-Anatolian Region From 3D Finite Element Models, pp , Friedemann Wenzel, Springer, Berlin, Heidelberg. Jackson, J., and D. McKenzie (1988), The relationship between plate motions and seismic moment tensors, and the rates of active deformation in the Mediterranean and Middle East, Geophys. J., 93, Kaneko, Y., Y. Fialko, D. T. Sandwell, X. Tong, and M. Furuya (2013), Interseismic deformation and creep along the central section of the North Anatolian Fault (Turkey): InSAR observations and implications for rateand-state friction properties, J. Geophys. Res. Solid Earth, 118, , doi: /2012jb Le Pichon, X. (1968), Sea-floor spreading and continental drift, J. Geophys. Res., 73(12), Le Pichon, X., and C. Kremer (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, , doi: /annurev-earth Marinkovic, P., and Y. Larsen (2013), Consequences of long-term ASAR local oscillator frequency decay - An empirical study of 10 years of data, European Space Agency, Living Planet Symposium (abstract), Edinburgh, U. K. McClusky, S., et al. (2000), Global Positioning System constaints on plate kinematics and dynamics in the eastern Mediterranean and Caucasus, J. Geophys. Res., 105(B3), McKenzie, D., and R. L. Parker (1967), The North Pacific: An example of tectonics on a sphere, Nature, 216(5122), Nocquet, J.-M. (2012), Present-day kinematics of the mediterranean: A comprehensive overview of GPS results, Tectonophysics, 579, , doi: /j.tecto Ozener, H., E. Arpart, S. Ergintav, A. Dogru, R. Cakmak, B. Turgut, and U. Dogan (2010), Kinematics of the eastern part of the North Anatolian Fault Zone, J. Geodyn., 49, , doi: /j.jog Ozeren, M. S., and W. E. Holt (2010), The dynamics of the eastern Mediterranean and eastern Turkey, Geophys. J. Int., 183, , doi: /j x x. Reilinger, R., S. McClusky, and P. Vernant (2006), GPS constraints on continental deformation in the Africa-Arabia-Eurasia continental collision zone and inplications for the dynamics of plate interactions, J. Geophys. Res., 111, B05411, doi: /2005jb Rosen, P. A., S. Hensley, G. Peltzer, and M. Simons (2004), Updated repeat orbit interferometry package released, Eos Trans. AGU, 85(5), 47. Savage, J. C., and W. H. Prescott (1978), Asthenosphere readjustment and the earthquake cycle, J. Geophys. Res., 83, Smith-Konter, B. R., D. T. Sandwell, and P. Shearer (2011), Locking depths estimated from geodesy and seismology along the San Andreas Fault System: Implications for seismic moment release, J. Geophys. Res., 116, B06401, doi: /2010jb Tatar, O., et al. (2012), Crustal deformation and kinematics of the Eastern Part of the North Anatolian Fault Zone (Turkey) from GPS measurements, Tectonophysics, , 55 62, doi: /j.tecto Thatcher, W. (2009), How the continents deform: The evidence from tectonic geodesy, Annu. Rev. Earth Planet. Sci., 17, , doi: /annurev.earth Uppala, S., P. Kȧllberg, A. Simmons, and U. Andrae (2005), The ERA-40 re-analysis, Q. J. R. Meteorol. Soc., 131,

7 Walters, R. J. (2013), Geodetic observation and modelling of continental deformation in Iran and Turkey, D.Phil. thesis, Univ. of Oxford, Oxford, U. K. Walters, R. J., R. J. Holley, B. Parsons, and T. J. Wright (2011), Interseismic strain accumulation across the North Anatolian Fault from Envisat InSAR measurements, Geophys. Res. Lett., 38, L05303, doi: /2010gl Wright, T., B. Parsons, and E. Fielding (2001), Measurement of interseismic strain accumulation across the North Anatolian Fault by satellite radar interferometry, Geophys. Res. Lett., 28(10), Wright, T., B. Parsons, P. C. England, and E. Fielding (2004), InSAR observations of low slip rates on the major faults of western Tibet, Science, 305(5681),