Continued Convergence After the Occurrence of a Slab Break‐Off: The Case of the Cyprian Arc

The detachment (i.e., break‐off) of down‐going subducting oceanic slabs is a major geodynamic event with far‐reaching consequences, one of which is the reduction of the slab pull force acting on the trailing plate. We investigate the motion of the Sinai Microplate where a recent (∼1 Myr ago) slab break‐off occurred along its sole converging plate boundary (Cyprian Arc) with the overriding Anatolia Microplate. Based on new bathymetric mapping, high‐resolution seismic reflection imaging, geodetic and earthquake data, we show that Sinai is actively moving in a northwest direction with respect to Nubia. Our results indicate that despite the recent slab break‐off, Sinai has and is still being pulled (or pushed) toward the overriding Anatolia Microplate. The continued convergence possibly occurs because of a persistent slab pull force, a suction force induced by the down‐going detached slab and/or by the upper mantle flow induced by the Afar Plume.


Introduction
Subduction of the oceanic lithosphere into the mantle exerts slab pull force on the trailing plate, the prime force that drives plate tectonic motions (Forsyth & Uyeda, 1975).In general, subduction may continue until the trailing passive margin and continental lithosphere enter the trench and some 5-15 Myr later, detachment of slab (i.e., break-off) occurs, and the subduction is terminated (Gün et al., 2021;Hafkenscheid et al., 2006;Wortel & Spakman, 2000).The termination of subduction has far-reaching geodynamic consequences, one of which is the initiation of changes in plate motions that may ignite a plate tectonic chain reaction (Gürer et al., 2022).Geological observations (e.g., McQuarrie et al., 2003) and numerical models (Duretz et al., 2011;van Hunen & Allen, 2011) suggest that convergence velocities between the subducting and overriding plates decrease once continental collision starts.However, contrasting predictions exist for the kinematic influence of slab break-offs, with some of the models predicting a complete halt in convergence velocities (although vertical rebound is initiated) (Duretz et al., 2011;Göğüs & Pysklywec, 2008;van Hunen & Allen, 2011), while others predict that subduction of the trailing plate should continue due to the suction forces induced by the sinking of the detached slab (Knight et al., 2021).The detachment of subducting slabs has long been observed by the identification of gaps in seismic activity or tomographic images (e.g., Isacks & Molnar, 1971;Wortel & Spakman, 2000), yet the commonly complex nature of the subduction plate boundaries makes it hard to identify and isolate the dynamic influence of slab break-off.
The Sinai Microplate, confined between the Anatolian, Arabian, and Nubian Plates (Figure 1), was formed during the Late Cenozoic fragmentation of northeast Africa and provides a unique opportunity to investigate how a recent slab break-off (Kounoudis et al., 2020;Portner et al., 2018) impacts plate kinematics.The Cyprian Arc that forms the northern edge of the microplate has been its sole confining convergence plate boundary since the creation of the microplate in the Miocene (Le Pichon & Francheteau, 1978).Tomographic imaging suggest that the leading oceanic lithosphere found under the Cyprian Arc detached only ∼1 Myr ago from the trailing Sinai Microplate (Kounoudis et al., 2020;Portner et al., 2018).Unlike the slab break-off imaged along the northern boundary of the Sinai Microplate, the northern Nubian Plate, west of the microplate, is still actively subducting northward beneath Anatolia and the Aegean Sea Microplates (Figure 1a).Thus, the kinematics of the Sinai Microplate allows for assessing the impact of a slab break-off and the relative importance of the different forces acting on the Sinai Microplate.Here we provide new marine geophysical and geodetic evidence, as well as earthquake observations, for the continued northwestward motion of the Sinai Microplate with respect to the Nubian Plate that indicates that Sinai has and is still being pulled (or pushed) toward the overriding Anatolia Microplate, despite the recent slab break-off.
The Sinai Microplate consists of a continental lithosphere stretched and thinned toward the west.The Levant Basin contains a thinned (moho depth of ∼20 km) continental crust (Figure 1b, Netzeband et al., 2006), that was shaped through several extensional episodes during the Permian, Triassic, and Early Jurassic times (Garfunkel, 2004).The microplate is bounded on the east by the sinistral Dead Sea transform fault system (Sinai-Arabia plate boundary), and on the north, it is bounded by the Cyprian Arc that accommodates converging motion relative to the Anatolian Microplate.The northern boundary consists of an arcuate depression (i.e., trench) where fragments of the stretched Levant continental lithosphere are actively underthrusting beneath Cyprus (Robertson, 1998).Tectonic reconstructions (McPhee & van Hinsbergen, 2019) suggested that the temporal transition from subduction of oceanic lithosphere to continental collision started when the northern African passive margin  (Bird, 2003) and the vectors indicate the current relative velocities of the plates with respect to the Eurasian Plate (Argus et al., 2011).The motion of the Sinai Microplate is calculated based on our Sinai-Arabia pole parameters combined with Arabia-Eurasia rotation.The dashed black line at latitude 37ºN marks the surface projection of the seismically imaged tear between the leading oceanic slab and the trailing Sinai Microplate.(b) Tectonic map of the Sinai Microplate.Brown vectors represent GPS-derived velocities of the Sinai Microplate with respect to a fixed Nubia Plate (Supporting Information S1 and Figure S1).The earthquake focal mechanisms for the Suez Rift and the Mediterranean Sea are taken from Jackson et al. (1988) and the USGS Catalog, respectively.The seaward edge of the continental-oceanic boundary (COB, dashed line) is taken from Granot (2016).The white polygon highlights the position of the investigated area shown in panel (c).ES, SR, and DSF stand for the Eratosthenes Seamount, Suez Rift, and the Dead Sea Fault, respectively.
entered the Cyprian trench in the Late Miocene (∼9 Ma).The subduction of continental material under Cyprus correlates with a tomographically imaged region of a northward-directed shallow sub-horizontal subduction having a relatively low seismic activity (Kounoudis et al., 2020;Portner et al., 2018).These seismic tomographic models also show that the shallow and flat slab is terminated in the north by a ∼200 km-deep narrow gap (marked by a dashed line in Figure 1a).This gap separates the flat trailing plate from a large vertically dipping block of a down-going slab.These geological as well as seismic tomography observations suggest that the down-going oceanic slab has detached from the trailing, most likely stretched continental, Sinai Microplate.Such slab tearing along the oceanic-continental transition zone at around 200 km depth has previously been predicted by numerical models (Gün et al., 2021;van Hunen & Allen, 2011).We note that although detached at depth from the leading oceanic slab, the tomographic images suggest that the shallow subducted continental lithosphere (i.e., <200 km depth) is still being attached to the subducting oceanic lithosphere found west of Cyprus (part of the Nubian Plate).Once slab break-off occurred, the velocity of the leading detached oceanic slab has probably accelerated to velocities on the order of ∼10 2 mm/yr (van der Wiel et al., 2024) thus the narrow gap (∼50 kmwide) between the detached and trailing lithosphere might have developed over a relatively short time period.The emergence of a recent uplift observed in Cyprus (∼800 m of vertical uplift during the past 1 Myr, Palamakumbura et al., 2016;Poole et al., 1990) further suggests that the slab break-off occurred sometime in the Pleistocene.
The Sinai Microplate is bounded in the south by the Gulf of Suez rift system.Structural data (fault trends and fault-slip data) indicate that since the creation of the rift in the Late Oligocene, it accommodated pure extensional motion in an N65ºE direction (Khalil & McClay, 2001).Toward the northwest, the offshore continuation of the rift is marked by a series of long (∼200 km) NW-SE-oriented pronounced seafloor escarpments forming a tectonic corridor marking the western boundary of the microplate (Figure 1b).This tectonic corridor seems to reactivate an ancient inherited basement faults (i.e., rift valey) (Aal et al., 2000).This rift points toward the mechanical boundary where the strong oceanic crust of the Herodotus Basin meets with the weak and stretched continental lithosphere of the Levant Basin (Granot, 2016).A seismic reflection study traced these NW-SE long lineaments deep into the sedimentary sequence, below the Messinian salt layer (Mascle et al., 2000), and concluded that they are the surface exposures of strike-slip faults.Based on morphological analysis of salt ridges, the motion along these offshore NW-SE faults was deduced to be of mostly dextral sense (Loncke et al., 2006;Mascle et al., 2000).Accordingly, these works have inferred that the Gulf of Suez also accommodates dextral plate motion.Although contradicting the geodetic and seismicity data from the Gulf of Suez (Figure 1b), this kinematic model suggests that the collision of Sinai with Anatolia led to the inversion of motion and emergence of dextral relative motion between Sinai and Nubia (i.e., along the tectonic corridor).Alternatively, based on analog modeling, Loncke et al. (2010) suggested that the dextral motion along these offshore faults could be explained by a northward gravity spreading of the Messinian salt layer against the Eratosthenes Seamount that acts as a buttress.This implies that the nature of activity along the NW-SE fault scarps is, in fact, caused by a thin-skinned motion that has no connection with the present-day plate kinematics of the eastern Mediterranean.Although clouded with uncertainties, knowing the sense and cause of the motion along the western boundary of the Sinai Microplate is critical in order to reconcile long-term geological observations with geodetic plate kinematic models.Furthermore, this knowledge will help to understand the geodynamic impact of the slab break-off: if the northward plate motion of Sinai has slowed down to a halt due to the slab break-off, then a dextral plate motion is expected to govern the western boundary of the microplate.Alternatively, if the offshore Sinai-Nubia motion is sinistral, then it indicates that the microplate is still being pulled (or pushed) toward the Anatolia Microplate, despite the detachment of the leading oceanic slab.

Methods
We conducted a marine geophysical survey in May 2020 aboard RV Bat Galim, aimed at surveying the primary fault straddling the eastern edge of the tectonic corridor that marks the western boundary of the Sinai Microplate (Figures 1 and 2).We collected multibeam bathymetric data (using a Kongsberg EM302 sonar) to map the seafloor in 30 m resolution and acquired seven two-dimensional high-resolution seismic reflection profiles, oriented at a high angle with respect to the fault and related structures.We used the Geo-Marine Survey Systems 150 m-long 48 channels streamer (Geo-Sense Ultra Hi-Res, 3.125 m spacing between channels) coupled with their Geo-Source 400 Sparker that supplies 6-kj pulsed power operating at 0.5-3 kHz.The streamer and source depth was 0.3 m.Data were acquired at a sampling interval of 0.1 msec, yielding sub-meter resolution with a penetration of about 150-200 m.Data processing included geometry, band-pass filtering, Crooked line common midpoint binning, and removal of bad coordinate traces.We then performed a velocity analysis based on picking of pre-stack semblance curves, normal move out correction, stacking, post-stack time migration, and amplitude gain.The seismic results, and the archival 6-channel seismic reflection data (Mascle et al., 2000), were then imported into Kingdom Suite software for interpretation.
We updated the Sinai-Nubia rotation parameters using geodetic data collected between 2001 and 2019 from 53 permanent GPS stations located at the Sinai, Nubian, and Arabian Plates and 21 additional stations in Anatolia and Eurasia Plates (see Supporting Information S1 for details).Finally, we compared our geodetic kinematic results with the previously published geodetic models (Kreemer et al., 2014;Mahmoud et al., 2005;Saleh & Becker, 2015;Wdowinski et al., 2004).

Results
The new bathymetric maps provide a 50 km-long detailed view of the investigated fault and its surroundings (Figure 2).In the south, the fault splits into a secondary fault that straddles northward (N13ºW) and dies out some 20 km north of the junction, whereas the main activity seems to focus along an N37ºW-oriented fault.This main branch (highlighted by black arrows in Figure 2a) comprises two subparallel fault segments spaced a few hundred meters apart and dip almost vertically (Figure 3).These fault segments also create ridges and troughs along strike and separate strata with varying thicknesses, all of which suggest that the main NW-SE fault accommodate predominantly strike-slip motion.At around latitude 33.15°, a series of morphological hills splay westward from the main fault.These hills have an asymmetric cross-sectional shape whereby their steep (9°±1°) flanks face toward the north-northeast, whereas their southern flanks have lower slope angles (4°±1°).The seismic reflection profiles (Figure 3) show that prominent faults truncate their northern flanks and that the sedimentary strata dip southward, suggesting that normal faulting caused gentle backward (i.e., southward) tilting.These faults may have also accommodated some horizontal displacement yet bathymetric indications of such motion are likely to be eroded and reworked by the sedimentation system active in this region.Thus, we refer to these faults as transtensional faults.Growth strata structures found within the ∼50 to ∼15 ms depth two way travel time indicate syntectonic deposition (see inset in Figure 3b).This observation suggests that a phase of faulting has been active during the past ∼1 Myr (assuming a constant sedimentation rate of ∼1.6 cm/kyr (Emeis et al., 1996) and a seismic wave velocity of 1,600 m/s).Evidence for earlier slip, if occurred, is located too deep for our data to resolve.Salt bodies, identified by chaotic seismic facies with velocity pull-ups at their steep boundaries, penetrate the strata at different depths and appear at shallower levels near the transtentional faults (Figure 3).Although these faults are not imaged well within the salt, we infer that the faults do not originate from the top of the salt diapirs (for instance, see Figure 3b) but are rooted deeper instead, favoring them as providing the conduits for upward salt propagation.Additionally, salt tectonic-driven normal faults are commonly associated with radially oriented faults and/or blind faults at the diapir top (e.g., Schattner et al., 2018), which we do not observe.Finally, since the observed prominent transtentional faults originate (i.e., splayed) from the main fault trace (Figure 2b), it is unlikely that they were formed due to salt motion but instead are secondary faults that were created under a regional ∼NNE-SSW (035º-215°± 10°) extensional horizontal stress regime.This faulting seems to have facilitated the upward motion of the salt (i.e., reactive diapirism due to extension, Hudec and Jackson (2007)).
Unlike the prominent transtentional faults that created the asymmetric hills, on the southwest side of the profiles (Figure 3), additional steep small-offset mostly southward-dipping faults cross-cut the upper sedimentary strata.These faults can be resolved down to the top of the salt body, where they seem to originate.They might have formed under the same conditions as the prominent transtentional faults that dip toward the north, but they might have formed instead, for instance, by salt motion.Nevertheless, due to their relatively small displacements, their existence does not violate our deduced direction of the regional stress field.
The stress regime inferred from the splayed transtensional faults suggests that the investigated 200 km-long NW-SE strike-slip fault accommodates a sinistral motion.This inference is supported by the geometry of structural and morphological features in the vicinity of the main strike-slip fault.The splayed asymmetric hills are truncated in the north by two morphological troughs oriented in a ∼NE-SW direction (highlighted by red arrows in Figure 2b).The seismic reflection profile shown in Figure 3a crosses the northern edge of one of these troughs and shows that channel-filled sediments were deposited within the trough.Interestingly, as the troughs meet with the main strikeslip fault, their orientation is sinistrally (counterclockwise) curved (Figure 2b).Curved landscape is commonly observed near strike-slip faults, for instance, near the San Andreas Fault (e.g., Gray et al., 2018) and the northern Dead Sea Fault (Goren et al., 2015), and was explained by distributed off-fault shear deformation.Thus, both the secondary transtensional faults and the curvature of the morphological troughs suggest that the investigated NW-SE strike-slip fault accommodates sinistral motion.Based on en echelon fold patterns, sinistral fault kinematics was previously suggested for the westernmost fault that bounds the offshore tectonic corridor (Figure 1b, Loncke et al., 2006).Collectively, these widespread observations suggest that within the eastern Mediterranean Sea, the Sinai Microplate has been moving in a northwest direction with respect to the Nubian Plate during the past ∼one million years.
Our kinematic inference that the Sinai Microplate has been moving northwestward with respect to the Nubian Plate generally agrees with our regional geodetic results and earthquake focal mechanisms.We perform a geodetic analysis (Supporting Information S1) that provides an updated Sinai-Nubia Euler rotation parameters (N29.44 ± 1.2°, E30.48 ± 2.2°, and Ω = 0.288 ± 0.009°deg/Myr).This rotation implies a northwest (i.e., sinistral) motion (∼1.7-1.9 mm/yr) with a possible slight compressional component of Sinai relative to Nubia at the study area (Figure 1b and Figure S1 in Supporting Information S1).The slow Sinai-Nubia relative motion accommodated in the area of the tectonic corridor in the Levant implies that relatively few earthquakes have been recorded there (Salamon et al., 2003).Unfortunately, due to their relatively small magnitudes, the focal mechanism solutions for these earthquakes are of limited quality.Importantly, reliable earthquake focal mechanism solutions (Mw > 5.5) from the southern edge of the Suez Rift (Figure 1b) and our GPS-based updated kinematic solution suggest that the Suez Rift still accommodates, albeit small (∼1-2 mm/yr), pure extensional Sinai-Nubia relative plate motion for most of its length.Overall, the relative Sinai-Nubia geodetic-based velocity vectors (Figure 1b) seem consistent with the Gulf of Suez earthquake focal mechanisms and our offshore geophysical data.

Discussion and Conclusions
Our marine geophysical analysis and the geodetically derived regional kinematic framework suggest that a northwestward offshore motion of the Sinai Microplate with respect to the Nubian Plate has prevailed during the last ∼one million years.This inference, in turn, suggests that the sense of motion along the plate boundary has  (Faccenna et al., 2013).Active converging plate boundaries are shown with green lines, where the age of slab break-offs, if occurred, are indicated (Hafkenscheid et al., 2006;McPhee & van Hinsbergen, 2019;Portner et al., 2018).Note that the green lines indicate the location of the plate boundaries and not the location of the slab tears.Inset is a schematic block diagram showing the geometry of the Cyprus Arc subduction zone (Kounoudis et al., 2020;Portner et al., 2018) and the horizontal tear that led to the break-off of the leading oceanic slab from the trailing Sinai Microplate.

Geophysical Research Letters
10.1029/2023GL108095 remained stable since the Early Miocene (Joffe & Garfunkel, 1987).The Dead Sea transform fault system that currently slips at ∼5 mm/yr (Hamiel & Piatibratova, 2021), similar to the long-term average slip rate that was accommodated over its entire life span since the Early Miocene (e.g., Freund et al., 1968;Nuriel et al., 2017), indicates that a relatively stable plate motion has also prevailed between Sinai and Arabia.
The Nubian Plate is actively being subducted under the Anatolia and Aegean Sea Microplates along the Hellenic Arc (DeMets et al., 2010;Figures 1a and 4).The Sinai Microplate moves in a northwestward direction with respect to Nubia, and is therefore currently converging with the overriding Anatolia Microplate.If the Nubian Plate had induced the force that drives the Sinai-Anatolia convergence by shearing along the tectonic corridor (i.e., leading to the clogging of the continental lithosphere under Cyprus), a Nubia-Sinai dextral motion would have been expected south of the Eratosthenes and along the Sinai-Nubia plate boundary (Figure 1b).The observed Sinai-Nubia sinistral relative motion thus indicates that the continental collision does not dominate the regional kinematics.
The ongoing convergence of Sinai with respect to Anatolia along the Cyprian Arc indicates that the force acting on the trailing Sinai Microplate may not have dissipated entirely, as predicted by some of the numerical simulations for a slab break-off occurrence (e.g., van Hunen & Allen, 2011).The slab pull force may persists due to the connection of the shallow subducting, probably continental, lithosphere with the oceanic lithosphere that is actively subducting under Anatolia, along the western Cyprian Arc (Inset in Figure 4; Granot, 2016;Portner et al., 2018).Alternatively, the Sinai-Anatolia convergence motion might be driven by suction forces induced by the down-going detached oceanic slab, as predicted by the numerical models of Knight et al. (2021).Alternatively, or in combination with, an external northward-directed push force may act on the microplate.Mantle plumes induce large-scale mantle flow that may facilitate such push forces (Cande & Stegman, 2011;van Hinsbergen et al., 2021).Indeed, mantle circulation modeling of the eastern Mediterranean region indicates that a large-scale northerly upper asthenospheric flow is fueled by the upwelling of mantle material at the Afar plume (Faccenna et al., 2013), located to the southeast of the microplate (Figure 4).Slab break-off has occurred along the Bitlis-Zagros collision plate boundary at around ∼12 Ma (Hafkenscheid et al., 2006), thus no suction forces are currently expected to be induced by the down-going northern Arabian detached slab.Thus, the push force induced by the Afar plume has been suggested to be the main force driving the northward motion of the Arabian Plate with respect to the Eurasian Plate.The Sinai Microplate, although located on the western edge of the region influenced by the Afar plume (Faccenna et al., 2013), might also be pushed northward, toward Anatolia, by the induced upper asthenospheric manle flow (Figure 4) and/or by drag force acting on the Dead Sea Fault plate boundary by the Arabian Plate.Additional upper mantle flows related to the complex subduction system (Göğüs & Pysklywec, 2008) may also contribute to the northward motion of Sinai.
We conclude that the post-break-off motion of the trailing continental Sinai Microplate along its sole converging plate boundary (Cyprian Arc) is fueled by some combination of slab pull and suction forces and upper mantle push force.This inference suggests that the detachment of subducting slabs may not immediately terminate the subduction process, which may continue with or without the addition of external forces.

Figure 1 .
Figure 1.(a) Tectonic overview map of the Levant and surrounding area.The plates are color-coded(Bird, 2003) and the vectors indicate the current relative velocities of the plates with respect to the Eurasian Plate(Argus et al., 2011).The motion of the Sinai Microplate is calculated based on our Sinai-Arabia pole parameters combined with Arabia-Eurasia rotation.The dashed black line at latitude 37ºN marks the surface projection of the seismically imaged tear between the leading oceanic slab and the trailing Sinai Microplate.(b) Tectonic map of the Sinai Microplate.Brown vectors represent GPS-derived velocities of the Sinai Microplate with respect to a fixed Nubia Plate (Supporting Information S1 and FigureS1).The earthquake focal mechanisms for the Suez Rift and the Mediterranean Sea are taken fromJackson et al. (1988) and the USGS Catalog, respectively.The seaward edge of the continental-oceanic boundary (COB, dashed line) is taken fromGranot (2016).The white polygon highlights the position of the investigated area shown in panel (c).ES, SR, and DSF stand for the Eratosthenes Seamount, Suez Rift, and the Dead Sea Fault, respectively.

Figure 2 .
Figure 2. Bathymeteric mapping and tectonic analysis.(a) Multibeam bathymetry map of the area surrounding the investigated fault highlighted by black arrows.(b) Detailed bathymetric map showing the sinistral NW-SE fault segments and the surrounding secondary faults and morphologies.Rectangular tabs mark the downthrown sides of transtensional faults.Gray lines highlight the locations of the seismic reflection profiles.The thick red lines mark the location of the profiles shown in Figure 3.The small red arrows mark the position where morphological features are sinistrally curved as they approach the main NW-SE strike-slip fault.The white rectangle highlights the position of one of these curved morphological features shown in the inset.The large arrowheads mark the inferred orientation of the minimum (σ 3 ) and maximum (σ 1 ) horizontal stresses.The arcs positioned in front of the arrowheads display a rough estimation of the uncertainty (±10°) related to the direction of the stress field.A similar uninterpreted bathymetric map is found in Figure S2 in Supporting Information S1.

Figure 3 .
Figure 3. Representative northeast-southwest high-resolution seismic reflection profiles that cross the splayed hills and main faults (see Figure 2 for locations).Black lines mark the position of faults.Insets in Profile a are close-ups of channel-filling structures, whereas the inset in Profile b shows a close-up of growth strata deposited in front of a transtensional fault.The upper limit of the salt bodies is highlighted with dashed lines, and general upward propagation directions are marked with gray arrows.The interpreted horizons (blue and yellow) marked on the two sides of the main faults at the bottom of Profile b highlight lateral variations in the thickness of the strata, suggesting possible horizontal displacement, typical of strike-slip faulting.The box at the top shows a line drawing of the NE-SW deep seismic reflection profile (P44) of Mascle et al. (2000).The black rectangular shown on the box marks the location of the two seismic profiles shown in Profiles a and (b).The location of profile P44 and the uninterpreted version of our seismic reflection profiles are found in Figures S2 and S3 in Supporting Information S1, respectively.

Figure 4 .
Figure 4. Schematic interpretation of the dynamics surrounding the Sinai Micropate.Red arrows indicate the upper asthenospheric mantle flow originated from the Afar plume(Faccenna et al., 2013).Active converging plate boundaries are shown with green lines, where the age of slab break-offs, if occurred, are indicated(Hafkenscheid et al., 2006;McPhee & van Hinsbergen, 2019;Portner et al., 2018).Note that the green lines indicate the location of the plate boundaries and not the location of the slab tears.Inset is a schematic block diagram showing the geometry of the Cyprus Arc subduction zone(Kounoudis et al., 2020;Portner et al., 2018) and the horizontal tear that led to the break-off of the leading oceanic slab from the trailing Sinai Microplate.