Arava

Abstract:

Timing of activity along the Yammuneh segment (Lebanon-Syria) of the Dead Sea Fault Zone and its northward continuation is still a subject of controversy. Our field structural analysis and observations on radar, Landsat and digital elevation model imagery of the Horns region (Syria) are of concern for Plio-Quaternary tectonics of the whole northern part of the Dead Sea Fault Zone. We show in this paper that the northern Dead Sea Fault Zone has remained active since the onset of the Homs basalts at c. 6 Ma. Continuing movement in Recent times is indicated by the occurrence of Quaternary pull-apart structures and offset of active ravines along the fault. The Horns basalts are related to the distinct oval-shaped Shin volcanic edifice, of which the long axis trends NW. The volcano was fed through NW-striking tension fractures, which now form dykes and volcanic ridges. These patterns are consistent with a NE-SW extension that occurred c. 6 Ma ago. The northwestern end of the Shin volcano is left-laterally displaced c. 20 km, yielding a c. 3.3 mm a^sup -1^ mean rate of relative movement between the Arabian and African plates. In the northern part of the Dead Sea Fault Zone, the overall trace of the main active Dead Sea Fault Zone is not a single transform but forms an irregular plate boundary composed of transform fault and collision belt segments. Fine-grained mylonite developed in the fault corridor may have favoured aseismic deformation in the Shin volcano area.

Keywords: Dead Sea Fault Zone, Syria, Lebanon, remote sensing, pull-apart structure.

The Dead Sea Fault Zone runs parallel to the Levant coast of the Eastern Mediterranean and forms the boundary between the African and Arabian plates (Fig. 1). Occurrence of Recent activity along the Yammuneh (Lebanon-Syria) segment and its northward continuation is still a subject of controversy. Several workers have maintained that this segment is poorly active and deduced that the plate boundary has shifted since c. 6 Ma to the Roum fault, which branches near Beirut and continues to the NW into the Mediterranean (Girdler 1990; Butler et al. 1997, 1998; Butler & Spencer 1999). This interpretation implies that the Africa-Arabia plate boundary runs almost across Beirut and connects with the Anatolia-Africa boundary at a triple junction located in the Mediterranean, east of Cyprus. In contrast, Khair et al. (2000) concluded from a study of earthquake data that the Yammuneh segment and its northward continuation have experienced strong historical seismicity. Meghraoui et al. (2003) have demonstrated the occurrence of historical large earthquakes north of the Horns basalts. In this thinking, the northern part of the Dead Sea Fault Zone would then be active until it reaches the Africa-Arabia-Eurasia triple junction near Maras.

The Pliocene Horns basalts (north of Horns, Syria), mapped by Dubertret (1962), are c. 150 km north of the junction of the Roum and Yammuneh faults (Figs 1 and 2). Lava flows crop out on each side of the Dead Sea Fault Zone. Butler et al. (1997) pointed out that no deformation can be observed in these basalts at locations close (100 m) to the Lebanese segment of the Yammuneh fault. They consequently interpreted the basalts as resting disconformably on the fault, and argued that no noticeable displacement had occurred along this fault since basalt emplacement. In contrast, Fleury et al. (1999) argued for a strong expression of the Dead Sea Fault Zone across the Horns basalts, especially when observed from satellite imagery, and concluded that the fault zone crosscuts the basalts.

The aim of this paper is to present new observations in the Horns basalts area. We believe that part of the motion (c. 20 km) along the Syrian-Lebanese fault segment of the Dead Sea Fault Zone occurred since the emplacement of the basalts. First we overview the study area. We then describe our observations from satellite imagery and field structural analysis. In the final section, we conclude that the fault segment is active and explain our view of the Dead Sea Fault Zone as a present-day plate boundary.

Regional framework

An overall structural map of the left-lateral Dead Sea Fault Zone is shown in Figure 1. Information for the Gulf of Aqaba is from Ben Avraham & Zoback (1992), for the Arava and Dead Sea segments from Garfunkel (1981) and Garfunkel & Ben-Avraham (1996), for the Jordan segment from Garfunkel et al. (1981) and Hurwitz et al. (2002), for the Yammuneh restraining bend from Ron (1987) and Walley (2001 ), for the El-Ghab basin from Kopp et al. (1999), and for the Karasu segment fromPerinçek & Cemen (1990), Chorowicz et al. (1994), Rojay et al. (2001) and Adiyaman & Chorowicz (2002). The Gulf of Aqaba-Arava-Dead Sea-Jordan Valley segment comprises pull-apart basin structures and its trend curves generally from N10°E in the south to N05°E in the Jordan Valley. The Yammuneh segment strikes N35°E. The El-Ghab segment, also comprising pull-apart structures, strikes north-south and connects with the N25°E-striking Karasu segment.

From historical seismic data Khair et al. (2000) concluded that the strongest earthquake activity occurs in two segments of the Dead Sea Fault Zone: the Gulf of Aqaba (Klinger et al. 1999) and the Yammuneh and Karasu fault zones. The seismicity in the northern segments of the Dead Sea Fault Zone, extending northwards from the Yammuneh fault, is not significantly less than that in the southern segments formed by the Jordan Valley and the regions further south (Salamon et al. 1996; Ambraseys & Jackson 1998; Al-Tarazi 1999).

However, several workers (Quennell 1984; Girdler 1990; Butler et al. 1997, 1998; Butler & Spencer 1999), using structural analysis, have considered that the northern segment has not significantly moved during Recent times. As a consequence, they conceived the present-day plate boundary to form an arc that continues north of the Jordan Valley through the Roum fault, then near Beirut, to finally meet the Africa-Anatolia plate boundary somewhere east of Cyprus. We have drawn in Figure 1 the undersea continuation of this arc, as inferred by Butler et al. (1997). This interpretation was recently contradicted by Fleury et al. (1999), who argued that the morphology, as revealed by remote sensing imagery, expresses the characteristics of recent activity. Accordingly, Griffiths et al. (2000) concluded that the Roum fault dies out south of Beirut and does not continue into the Mediterranean.

Cenozoic volcanic occurrences along the Dead Sea Fault Zone, forming the 'Levantine Volcanic Province', are Miocene to Holocene in age (Mor 1993), with a gap in the activity between 16 and 8.5 Ma (Mouty et al. 1992). They are mainly composed of flood-type volcanic rocks distributed over three distinct regions: (1) the Harrat Ash Shaam plateau; (2) the region from the Horns basalts to the Karasu valley; (3) the Arabian platform and southern part of the Bitlis belt (e.g. Karacadag volcano).

The Harrat Ash Shaam basalt eruptions occurred in three episodes: at 26-22 Ma, 18-13 Ma and 7 to

The Horns basalts are dated at 6.5-2.0 Ma (Rukieh 1991; Mouty et al. 1992; Rukieh et al. 1994; Sharkov et al. 1994, 1998; Butler et al. 1997; Butler & Spencer 1999). In the Ghab basin area and east of it, the ages of volcanic rocks range from 2.0 to 1.1 Ma (Heimann et al. 1998). Eruptions were related to subvertical tension fractures accompanying the transcurrent movement along this part of the Dead Sea Fault Zone, and originated from adiabatic partial melting in the lithosphere only east of the Dead Sea Fault Zone, where the Afro-Arabian margin of the Eastern Mediterranean is thick enough for melting to occur (Polat et al. 1997; Adiyaman & Chorowicz 2002).

In the Karasu valley and vicinity, ages vary from 1.6 to 0.05 Ma (Capan et al. 1987; Heimann et al. 1998; Rojay et al. 2001; Yurtmen et al. 2002). The volcanism in this area depends mainly on extension related to lateral escape of Anatolia (Yürür & Chorowicz 1998). The Karacadag volcano is dated at 1.9-0.1 Ma (Notsu et al. 1995).

The Horns basalts lie on both sides of a N05°E-striking segment of the Dead Sea Fault Zone (Fig. 2). To the north, the fault trace steps over to the left, allowing the opening of the ElGhab pull-apart basin (Kopp et al. 1999). To the south, the fault connects with the N35°E-striking Yammuneh fault, forming a restraining bend associated with the transpressive structures of the Lebanon, Anti-Lebanon and Palmyrides Mountains. Further to the south, the N05°E-striking Jordan segment of the Dead Sea Fault Zone comprises the Hula (Zilberman et al. 2000) and Sea of Galilee (Hurwitz et al. 2002) pull-apart basins.

The Horns basalts, up to 850 m thick, are mainly composed of thin (1 m) to thick (13 m) lava flows, with intercalated pyroclastic deposits and a few red laterite levels (Rukieh et al. 1994). On the eastern side of the Dead Sea fault, they rest disconformably upon Aptian-Albian and Cenomanian carbonates (Fig. 3). In the west, they cover late Jurassic limestone and early to late Turonian carbonate (Ponikarov et al. 1963a, b). This difference in basement rocks under the basalts across the fault indicates that the western compartment was already uplifted when eruptions occurred, and consequently a post-Turonian (Miocene?) fault existed at this time. This is consistent with the observation made by Butler et al. (1997) and Butler & Spencer (1999) of the Homs basalts unconformably lying against a palaeo-fault scarp (site d in Fig. 4).

Our evidence of tectonics affecting the Horns basalts was obtained from remote sensing imagery analysis and complementary field observations. Fault activity is generally well expressed by the geomorphology, which can be best observed from radar imagery, digital elevation models and the IR band of the Landsat Thematic Mapper (TM) sensor. The radar image of Figure 5 was produced from a view of the European Remote Sensing satellite (ERS) Synthetic Aperture Radar (SAR) instrument. The DEM at 30 m pixel of Figure 6 is derived from the correlation of a Terra Aster stereoscopic couple. A detailed view of the fault across the Horns basalts in the visible-IR range of the Landsat TM is shown in Figure 7.

Observations

Satellite imagery and DEM data

ERS SAR image. The most striking feature observed in the ERS SAR image (Fig. 5) is the Shin hill, north of Horns, forming a distinct ovoid elongate landform that is typical of a volcanic edifice. Other surfaces that are, according to the geological map (Fig. 3), covered with thin lava flows and tephras do not show up in the image. We consider that the break in slope at the foot of the Shin hill and around it represents the base of the edifice, forming an ovoid shape with the main axis trending N135°E. The main axis of an elongate volcano generally is superimposed on a zone of tension fractures, which strike at right angles to the extension (e.g. Chorowicz et al. 1997; Dhont et al. 1998). The Shin volcano is consequently probably related to a group of major tension fractures that opened as a result of N45°E-trending extension. On the top, several parallel volcanic ridges trend N135°E. They correspond to fissures that fed the volcano. More than 100 NW-striking dykes, up to 3 m thick, have been observed in the field by Rukieh et al. (1994). These observations indicate that local N45°E-trending extension occurred during volcanic activity about 6 Ma ago.

A salient fact is that this volcano is sharply cut at its northwestern end by the Dead Sea Fault Zone, suggesting that a missing part has been displaced (Fig. 5). The missing part should be in a similar state of erosion and should therefore also show up by its morphology. In the radar imagery, two hills are large enough to deserve examination. Hill A is made of Jurassic limestone and basal Cretaceous rocks, with a thin Horns basalts cover (Fig. 3). This cannot be a significant part of the original Shin volcano. Hill B (Figs 3 and 5), with the Krak des Chevaliers at its highest point, corresponds to a deeply eroded thick volcanic pile, with short valleys and a sharp crest line. Its half-round shapes suggest an ancient half-crater or caldera. The base contour of the western slope of hill B is round, is included in the basaltic outcrops, and can therefore be taken to represent the westernmost end of the former Shin volcano (Fig. 5) before it was cut and displaced by the Dead Sea Fault Zone. The semicircle formed by hill slopes drawn in Figures 3 and 5 perfectly fits with the main volcano if it is displaced at least 16km northward.

In the south of the image, the Yammuneh fault turns from N35°E to N05°E. A 1.8km long rhomb-shaped pull-apart basin was formed here, filled with Quaternary deposits (PAl in Fig. 5). Further north, the offset of the northern ends of the Western Lebanon and Eastern Lebanon Mountains is c. 1 1 km, but the feet of the mountains are not well constrained in the radar image. The trace of the Dead Sea Fault steps over to the left when penetrating the southern border of the Bokaiah basin, and continues northward along the bottom of the scarp of the Krak des Chevaliers. We infer from this pattern that the Bokaiah basin originated from a pull-apart structure (PA2 in Fig. 5), but the plain is wider than a simple pull-apart basin because it increased in width by collapse along the bounding faults. Quaternary sedimentary fill testifies to recent subsidence of the basin. Along the valley cutting the northwestern part of the Shin volcano, a left-stepping offset of the main fault trace delineates a rhombshaped pattern that is not a Quaternary sedimentary fill but corresponds to already emplaced basalt rocks that subsided during the opening of a pull-apart structure (PA3 in Fig. 5). Another left-stepping relay of the Dead Sea fault (PA4 in Fig. 5) can be observed on the radar image at the northern boundary of the Shin volcano, together with round hills corresponding to small Quaternary volcanoes (Rukieh et al. 1994). The shape of this structure is consistent with that of a pull-apart structure, with its opening being responsible for ascent of magma instead of collapse, which would have been responsible for the formation of a subsiding Quaternary basin.

The pull-apart structures along this segment of the Dead Sea fault (PAl to PA4, Fig. 5) clearly originated from sinistral shearing, but different types of structures can be distinguished. PAl and PA2 are basins filled with Quaternary sediments. PA3 corresponds to collapsed Horns basalts, forming a low-lying area. PA4 is filled with Quaternary volcanic material.

Some slopes are steep and sharply incise ancient surfaces of ablation, as is the case south of PA2 and PA4, indicating active erosion. Local erosion expresses active local uplifts characterizing deformation along strike-slip faults.

Terra Aster DEM image. The digital elevation model of Figure 6 results from the correlation of two Terra Aster scenes. Blank areas (in grey) occur where clouds blanket the topography. Similar to the radar image (Fig. 5), this shadowed and coloured DEM image shows the main Dead Sea fault truncating the Shin volcano. The PAl pull-apart basin is well delineated (Fig. 6). The left-stepping offset of the main fault across the Bokaiah Quaternary basin is obvious, and consequently the inferred PA2 pull-apart is justified. The PA3 pull-apart structure and related offset of the main fault is partly observed.

The chief interest of this DEM image is that it displays the steep slopes that sharply cut erosional surfaces, south of PA2 and in the area of PA3. A narrow ravine cuts one of the scarps (5 in Fig. 6b). This active erosion expresses local uplift that is typical of an active strike-slip fault. The Krak des Chevaliers is located at the boundary of a distinct semi-circular valley having the morphology of an ancient crater or caldera. Considering that this was the northwestern end of the Shin volcano 6 Ma ago, we compute that the axis of the edifice on the eastern side of the Dead Sea Fault has been left-laterally displaced over c. 20 km. This value, based on the well-defined axis of the former Shin volcano, is better documented than the c. 16km offset obtained from displacement of the foot of the volcano observed in the radar image (Fig. 5).

The DEM image is particularly efficient in displaying variations in relative elevations. It shows that the northern borders of the Western and Eastern Lebanon Mountains (D-D' in Fig. 5) are shifted c. 11 km. This is consistent with the estimate of Butler & Spencer (1999).

Landsat TM image. We have produced, from an extract of a Landsat TM view, a colour composite image made from bands 2, 4 and 5, respectively displayed in blue, green and red (Fig. 7). Thick chlorophyllian vegetation is then shown in green as a result of high reflectance values in band 4 (near-IR). The predominance of the IR bands (bands 4 and 5) gives a better contrast between surfaces that face towards or away from solar illumination, allowing good expression of the morphology.

The image expresses the main fault traces and some of the volcanic ridges along the crest of the Shin volcano. Their N 135°E strike is consistent with a N45°E extension. The footline of the Shin volcano cannot be delineated in this type of imagery. Near site 8, late Jurassic limestone and dolomite are shown in white.

The main outcome of this image is the mapping of the drainage network that develops west of the Shin volcano. On the eastern slopes of the main N170°E0 -trending structural valley (Fl in Fig. 7), the active ravines downgrade westward but they tend to turn left before reaching the bottom, an indication of leftlateral active drag folding occurring near the fault. The ravines of the western valley slopes are composed of two well-defined parts, with a pronounced change in trend that expresses a fault (f in Fig. 7) running along the main axis of the pull-apart structure (PA3 in Fig. 7). Between faults f and Fl, the ravine segment has an S shape, showing that the compartment is subject to high strain, consistent with active left-lateral slip along the bounding faults.

Field structural analysis

Cross-section. The sites of our field observations are concentrated along the valley of the Dead Sea Fault Zone, where it is incised in the Horns basalts (Fig. 7). A field cross-section near sites 1 and 8 (Fig. 7) testifies to transpressional folds involving the basalt (Fig. 8). Fault Fl juxtaposes Jurassic dolomitic limestone and the Pliocene basalt. The volcanic layers have a pronounced (50°) dip to the east, as shown by the attitude of red and yellow palaeo-soils, which indicates deformation post-dating their formation. However, there is no cataclastic deformation in the basaltic layers, even only a few metres from fault Fl. The carbonate material is cataclastic, is cut by faults f and F2 (Fig. 7), and forms folds with axes parallel to the fault zone.

To the south, at site 5, a deep, young, straight ravine (Figs 6 and 7) cuts the anticline, which involves at this place the basalts overlying the Jurassic layers. Antecedence of the ravine explains why it was not affected by folding. Hence, we conclude that the ravine deepened together with active folding and related uplift. Faults and mylonite. To the south of site 5, the basalts crop out continuously in the valley. As pointed out by Butler et al. (1997), the fractures in the Horns basalts are not numerous, even close to the Dead Sea Fault Zone (within tens of metres of it). However, we found and analysed several faults that clearly affect the basalts. Our structural analysis consisted of measurements of tension fractures and the orientation and sense-of-movement of striated fault planes, plotted on Schmidt diagrams (Fig. 9). Emphasis has been placed on striations directly observed on the major mapped fault planes, on which the main part of the displacements occurred. We have also taken into account striations observed on smaller faults paralleling the nearby major fault, assuming that in a given local stress field parallel faults have the same mechanism, for a given tectonic phase. We have plotted the plunge of striations on the trace of the main faults indicated by bold lines, with symbols indicating the relative sense of movement. These symbols have been copied from the Schmidt diagram plots and pasted onto the map of Figure 7. This presentation of the data shows the horizontal component of the relative movement of two blocks separated by a given fault.

At site 1 (Fig. 9, location in Fig. 7) a set of faults cuts the basalts (example A in Fig. 10), with subhorizontal striations on the slickensides. The main fault, striking almost parallel to the plate boundary (bold line in diagram for site 1, Fig. 9), has horizontal left-lateral displacement as indicated by subhorizontal striations, with associated small, penetrative Riedel fractures also having strike-slip geometry.

At site 2 (Fig. 9, location in Fig. 7), the basalt flows are mylonitic only in places, the rocks there being composed of cataclastic elements of centimetre to metre size. Distinct faults crosscut the mylonitic basalts (Fig. 1Oc). A fault that lies parallel to the main N10°E-striking mapped fault has left-lateral obliqueslip offset with reverse component. Several associated faults have subhorizontal striations.

At site 3 (Fig. 9, location in Fig. 7), a fine-grained (millimetre size), flour-like, mylonitic rock forms the main mapped fault. It crosscuts a coarser mylonite composed of metre- to centimetresized clasts. The 2-5 m thick gouge zone is squeezed between two subvertical planes striking approximately north (Fig. 1Ob). Smaller east-striking faults, mainly with reverse throw component, indicate N160°E shortening. Subhorizontal striations and typical steps on a main mapped subvertical N15°E-striking plane indicate left-lateral strike-slip movement.

At site 7 (Fig. 9, location in Fig. 7), fault f is subvertical, with oblique-slip left-lateral striations. The associated faults can be considered as synthetic Riedel faults in a sinistral shearing environment. Two of them have subhorizontal striations.

Site 6 (Fig. 9), located at the southern end of the PA3 pullapart structure (Fig. 7), differs from the previous sites because it shows a predominance of extension. The faulted basalt outcrops are well exposed along a recently dug canal. Main normal-throw components are documented both by steps on the fault slickensides and by downthrown offsets of basaltic layers. In places, younger basaltic layers disconformably cover the faults affecting the basalts, showing that the deformation and the volcanism are partly coeval. The striations plunge in almost all directions, indicating that gravity collapse occurred in places. We have also observed small (5-10 cm wide) open tension fractures, filled with calcite. One of them is covered by deposits of an ancient Quaternary terrace; three of them do not affect the overlying layers; two others form tiny symmetrical graben structures (widths 25 cm and 3 m, with vertical throw of the bounded faults of 3 m and 2 m, respectively). This confirms that the tension fractures were formed during the volcanic activity. The main directions of extension deduced from orientations of tension fractures range from N15°E to N40°E this is consistent with the N45°E extension trend obtained from mapping of the strike ridges in the Shin volcano. We conclude that the syneruptive structures at site 6 correspond to local gravity collapse, which was related either to the opening of the tension fractures that constituted routes for the magma, or to the onset of extensional strain at the southern end of the PA3 pull-apart structure. Strikeslip movement occurred later along a north-striking fault, dipping W75°, where subhorizontal striations overprint older dip-slip striations (bold arrows and line on the north-striking fault at site 6 of Fig. 9).

Discussion and conclusions

Our structural analysis and observations show that the c. 6 Ma Horns basalts underwent left-lateral strike-slip deformation as a result of motion along the Dead Sea fault. Layers dip 50° at places; they are cut mainly by roughly north-striking faults, which have exposures of several hundreds of metres in length in the well-defined fault zone, and have distinct subhorizontaL striations associated with left-lateral movement indicators. Basaltic rocks have been transformed only in places along the plate boundary into mylonitcs and fine-grained gouge zones, but the carbonate basement rock is more systematically mylonitic, a fact that indicates that carbonates can be more easily fragmented than basalts in proximity to the Dead Sea Fault Zone. Still-continuing movements are indicated by the occurrence of several pull-apart basins filled with Quaternary sediments and by deformation of east-west-trending small (recent) active ravines. The slopes are very steep in places, forming scarps.

The Horns basalts are associated with a distinct volcanic edifice (the Shin volcano), which has an oval shape elongate in a N 1350E direction and was fed by fissures having the same strike. This edifice is crosscut at its northwestern termination by the Dead Sea Fault Zone, and part of it (Krak des Chevaliers, Figs 5 and 6) is left-laterally displaced c. 20 km, yielding a mean rate of relative movement of c. 3.3 mm a^sup -1^ since 6 Ma in this region. From displacement of volcanic outcrops in the same area, Garfunkel et al. (1981) estimated 8km left-lateral offset along the Dead Sea Fault Zone, but they did not observe the shape of the volcano, whose major axis yields a more accurate value. Our slip rate estimate is also to be compared with that of 5 mm a^sup -1^ for the late Pleistocene in the northern segment of the Dead Sea Fault Zone (Radwan et al. 1992), 4.1 mm a^sup -1^ for the Quaternary in the Karasu segment (Rojay et al. 2001) and 6 mm a^sup -1^ for the past 20 ka in the southern segment (Zak & Freund 1966; Freund et al. 1970; Garfunkel & Bartov 1977; Ben-Manahem 1981). Our estimate of the displacement of the Mesozoic basement is about 17.5 km, taking into consideration the boundary between the Jurassic and the Cretaceous units from point B in the western compartment (Fig. 3) to the latitude of point A in the eastern compartment.

The main observation that led to the concept of inactivity of the northern segment of the Yammuneh fault since Miocene time, best described by Butler & Spencer (1999), is the Horns basalt resting unconformably on unstructured carbonate gouge of the 'calcaires du Sannine' (Albian-Cenomanian). However, at the very place where this unconformity lies in the western compartment (location d in Fig. 4), the geological map of Guerre (1968) indicates that the boundary of the basalts turns west and does not follow the fault in two small areas. At location 7, the main fault is almost 1OO m further to the east at the bottom of Wadi Chadra valley. The Dead Sea Fault Zone and related Wadi Chadra valley are clearly straight here, as shown by various geological maps, the DEM and satellite imagery. The palaeo-fault probably already existed before eruption of the Horns basalts, for they disconformably cover Jurassic and Cretaceous rocks as well. It is therefore necessary to conclude that the post-Pliocene movements have reactivated this palaeo-fault zone.

The pull-apart structures in the study area are clearly of different types. PAl and PA2 are basins filled with Quaternary sediments. PA3 is collapsed Horns basalts, forming a low-lying area. PA4 is filled by Quaternary volcanic rocks. Lengths of the pull-apart structures (PAl to PA4, Fig. 5) yield maximum offsets for the studied segment of the Dead Sea Fault Zone. These are, from PAl to PA4, 1.8 km, 14km, 5.5 km and 1.1 km. Assuming a mean rate of 3.3 mm a^sup -1^ since 6 Ma, maximum ages for the opening of these pull-apart structures are 0.55 Ma for PAl, 4.25Ma for PA2, 1.6Ma for PA3 and 0.33 Ma for PA4. This indicates that changes in the geometry of the deformation may occur inside the fault zone, as suggested by Heimann & Ron (1993) for the Hula basin area located further south. An explanation for the variety of sizes and possibly ages of initiation of pull-apart structures along the major strike-slip fault zone is beyond the scope of the present study. The brittle deformation is likely to have been complex in space and time in the heterogeneous, thinly loaded, uppermost crustal part of a lithospheric strike-slip fault. It generally constitutes in cross-section a flower structure that is more straight and throughgoing at depth. We observed in the field that the fault zone comprises at each site several parallel strike-slip faults that do not all have the same importance (Figs 7-10). Locally and within a limited time interval, a given strike-slip fault may be more or less active than the others in the fault zone.

Our analysis shows that the part of the Dead Sea Fault Zone that lies north of the Roum-Yammuneh branching has been active during the last 6 Ma. North of the Hula basin, relative plate motion is shared between diverging fault splays forming a braided strike-slip system (Walley 1998). The Yammuneh fault and its northern continuation have absorbed the largest part of the movement, whereas the other structures, including the Roum fault, underwent smaller deformation. This view of still-continuing movement north of the Yammuneh fault is consistent with an 80 km left-lateral displacement along the Dead Sea Fault Zone since the Miocene in northern Syria, inferred from the offset of the front of the ophiolitic nappes obducted during the late Maastrichtian (AI-Maleh et al. 1992).

As a consequence, the plate boundary between Africa and Arabia north of the Hula basin has not shifted since 6 Ma from the Yammuneh fault to the Eastern Mediterranean area. Strong seismicity is still to be expected along the Yammuneh fault and its northward continuation to the Maras triple junction. However, the fine-grained mylonite corridor that developed inside the contact may favour aseismic movements. Salamon et al. (1996) have shown that the seismic efficiency of the Dead Sea Fault Zone is very low (about 7%), stressing the role of aseismic deformation. It is now clear that the present-day plate boundary does not run across the city of Beirut. However, activity of the Roum fault and in the Lebanon-Anti-Lebanon should be considered.

From the up-to-date scheme shown in Figure 1, we consider that the main active Dead Sea Fault Zone, based on faults following arcs of concentric circles (Quennell 1983) can be divided into four segments, ending to the north at the Maras triple junction (Fig. 11). Two of these segments correspond to concentric transform zones (the Dead Sea and Al-Ghab transforms), each consisting of a series of strike-slip faults and pullapart structures. The other two correspond to compressional fold and fault belts (the Lebanon-Anti-Lebanon-Palmyrides and the Amanos belts). The concentric circles we have drawn have their centre at 32°00'N, 15°30'E. They are consistent with the presentday movements shown by GPS data (McClusky et al. 2000). This overall scheme indicates an irregular and composite plate boundary that cannot be reduced to a single Dead Sea transform line.

This work is the result of co-operation between, on the one side, the GORS (General Organization of Remote Sensing) and the University of Damascus and, on the other side, the University of Paris 6 (Université Pierre-et-Marie Curie) and the UPPA (Université de Pau et des Pays de l'Adour). The project is supported by the French Embassy in Damascus and the French Ministry of Foreign Affairs. We are most grateful to R. Butler and Z. Garfunkel for helpful comments and discussions during the review process.

References

ADIYAMAN, O. & CHOROWICZ, J. 2002. Late Cenozoic tectonics and volcanism in the northwestern corner of the Arabian plate: a consequence of the strike-slip Dead Sea Fault Zone and lateral escape of Anatolia. Journal oj Vblcanology and Geothermal Research, \ 17, 327-345.

AL-MALEH, K., DELAUNI;, M., MOUTY, M. & PARROT, J.-F. 1992. Relations du front de la nappe ophiolitique du nord-ouest syrien avec son substratum de part et d'autre de la faille du levant: Baër-Bassit, Kurd Dagh. Comptes Rendus de l'Académie des Sciences, Série II, 314, 1195-1202.

AL-TARAZI, E.A. 1999. Regional seismic hazard study for the eastern Mediterranean (Iran s-Jordan, Levant and Antakia) and Sinai region. Journal of African Earth Sciences, 28, 743-750.

AMBRASEYS, N.N. & JACKSON, J.A. 1998. Faulting associated with historical and recent earthquakes in the Eastern Mediterranean region. Geophysical Journal International. 133, 390-406.

BEN AVRAHAM, Z & ZOBACK, M.D. 1992. Transform-normal extension and asymmetric basins: an alternative to pull-apart models. Geology, 20, 423-426.

BEN-MANAHEM, A. 1981. Variation of slip and creep along the Levant Rift over the past 4500 years. Tectonophysics, 80, 183-197.

BILAL, A. & TOURET, J.L.R. 2001. Les enclaves du volcanisme récent du rift syrien. Bulletin de la Société Géologique de France, 172, 3-16.

BUTLER, R.W. & SPENCER, S. 1999. Landscape evolution and the preservation of tectonic landforms along the northern Yammouneh Fault, Lebanon. In: SMITH, B.J., WHALLRY, W.B. & WARKE, P.A. (eds) Uplift. Erosion and Stability: Perspectives on Long-term Landscape Development. Geological Society, London, Special Publications, 162, 143-156.

BUTLER, R.W., SPENCER, S. & GRIFFITHS, H.M. 1997. Transcurrent fault activity on the Dead Sea Transform in Lebanon and its implications for plate tectonics and seismic hazard. Journal of the Geological Society, London, 154, 757-760.

BUTLER, R.W., SPENCER. S. & GRIFFITHS, H.M. 1998. The structural response to evolving plate kinematics during transpression evolution of the Lebanese restraining bend of the Dead Sea Transform. In: HOLDSWORTH, R.E., STRACHAN, R.A. & DEWEY, J.F. (eds) Continental Transpressional and Transtensional Tectonics. Geological Society, London, Special Publications, 135, 81-106.

CAPAN, U.Z., VIDAL, P. & CANTAGREL, J.M. 1987. K-Ar, Nd, Sr and Pd isotopic study of Quaternary volcanism in Karasu Valley (Hatay), N-end of Dead Sea Rift zone in SE Turkey. Yerbilimleri Bulletin of Earth Sciences. Hacettepe University, Ankara, 14, 165-178.

CHOROWICZ, J., LUXEY, P., LYBERIS. N., CARVALHO, J., PARROT, J.F., YURUR. T. & GUNDOGDU, N. 1994. The Maras Triple Junction (Southern Turkey) based on digital elevation model and satellite imagery interpretation. Journal of Geophysical Research, 99, 20225-20242.

CHOROWICZ, J., BARDINTZEFF, IM., RASAMIMANANA, G., CHOTIN, P., THOUIN, C. & RUDANT, J.-P. 1997. An approach using SAR ERS images to relate extension fractures to volcanic vents: examples from Iceland and Madagascar. Teclonophysics, 271, 263-283.

DHONT. D-, CHOROWICZ, J., YÜRÜR, T., FROGER, J.-L., KÞSE, O. & GÜNDOGDU, N. 1998. Emplacement of volcanic vents and geodynamics of Central Anatolia, Turkey. Journal of Volcanology and Geothermal Research, 85, 33-54.

DUBERTRET, L. 1962. Carte géologique du Liban, Syrie el bordure de pays voisins, 1:1 000000. Muséum d'Histoire Naturelle, Paris.

FLEURY, J., CHOROWICZ, J. & SOMMA, J. 1999. Discussion on transcurrent fault activity of the Dead Sea Transform in Lebanon and its implications for plate tectonics and seismic hazards. Journal of the Geological Society, London, 156,1243-1248.

FREUND, R., GARFUNKEL, Z. & ZAK, I. 1970. The shear along the Dead Sea Rift. Philosophical Transactions of the Royal Society of London, 267, 107-130.

GARFUNKEL, Z. 1981. Internal structure of the Dead Sea leaky transform (rift) in relation to plate kinematics. Tectonophysics, 80, 81-108.

GARFUNKEL, Z. & BARTOV, Y. 1977. The tectonics of the Suez rift. Geological Survey oj Israel Bulletin, 71, 44.

GARFUNKEL, Z. & BEN-AVRAHAM, Z. 1996. The structure of the Dead Sea basin. Tectonophysics, 266, 155-176.

GARFUNKEL, Z., ZAK, I. & FREUND. R. 1981. Active faulting in the Dead Sea Rift. Tectonophysics, 80, 1-26.

GIRDLER, R.W. 1990. The Dead Sea transform system. Tectonophysics, 180, 1-13.

GRIFFITHS, H.M., CLARK, R.A., THORP, K.M. & SPENCER, S. 2000. Strain accommodation at the lateral margin of an active transpressive zone: geological and seismological evidence from the Lebanese restraining bend. Journal of the Geological Society, London, 157, 289-302.

GUERRE. A. 1968. Carte géologique de la République Libanaise, 1:50000, TeIlKalakh. Direction des Affaires Géographiques, Beirut.

GUERRE, A. 1975. Carte géologique de la République Libanaise, 1:50000, Hamidiyé-Halba. Direction des Affaires Géographiques, Beirut.

HEIMANN, A. & RON, H. 1993. Geometric changes of plate boundaries along part of the northern Dead Sea transform: geochronologic and paleomagnetic evidence. Tectonics, 12,477^)91.

HEIMANN, A., ROJAY, B. & TOPRAK, V. 1998. Neotectonic characteristics of the Karasu fault zone, northern continuation of Dead Sea Transform in Anatolia (Turkey). In: Third International Turkish Geological Symposium, Ankara, 99.

HURWITZ. S., GARFUNKEL, Z., BEN-GAI, Y., REZNIKOV, M., ROTSTEIN, Y. & GVIRTZMAN, H. 2002. The tectonic framework of a complex pull-apart basin: seismic reflection observations in the Sea of Galilee, Dead Sea Transform. Tectonophysics, 359, 289-306.

ILANI. S., HARLAVAN. Y. & TARAWNEH, K. ET AI.. 2001. New K-Ar ages of basalts from the Harrat Ash Shaam volcanic field in Jordan: implications for the span and duration of the upper-mantle upwelling beneath the western Arabian plate. Geological Society of America Bulletin, 29, 171-174.

KHAIR. K., KARAKAISIS, G.F. & PAPADIMITRIOU, E.E. 2000. Seismic zonation of the Dead Sea Transform Fault area. Annali di Geojisica, 43(1), 61-79.

KLINGER, Y.. RIVERA, L., HAESSLER. H. & MAURIN, J.-C. 1999. Active faulting in the Gulf of Aqaba: new knowledge from the Mw 7.3 earthquake of 22 November 1995. Bulletin of the Seismological Society of America, 89, 1025-1036.

KOPP. M.P., ADZHAMYAN, ZH., IL'YAS, K., FAKIANI, F. & KHAFEZ, A. 1999. Mechanism of formation of the El Ghab wrench graben (Syria) and the Levant Transform fault propagation. Geotectonics, 33, 408^122.

McCLUSKY, S., BALASSANIAN, S. & BARKA, A. ET AL. 2000. Global positioning system constraints on plate kinematics and dynamics in the eastern Mediterranean and Caucasus. Journal of Geophysical Research, 105, 5695-5719.

MEGHRAOUI, M., GOMEZ, F. & SBEINATI, R. ET AL. 2003. Evidence for 830 years of seismic quiescence from palaeoseismology, archaeoseismology and historical seismicity along the Dead Sea fault in Syria. Earth and Planetary Science Letters, 210, 35-52.

MOR, D. 1993. A timetable for the Levant Volcanic Province, according to K-Ar dating in the Golan Heights, Israel. Journal of African Earth Sciences, 16, 223-234.

MOUTY, M.. DELALOYE. M., FONTIGNIE. D., PISKIN. O. & WAGNER, J.-J. 1992. The volcanic activity in Syria and Lebanon between Jurassic and Actual. Schweizerische Mineralogische und Petrographische Mitteilungen, 72, 91-105.

NOTSU. K., FIJITANI. T., Ui, T., MATSUDA, J. & ERCAN. T. 1995. Geochemical features of collisional-related volcanic rocks in central and eastern Anatolia, Turkey. Journal of Volcanology and Geolhermal Research, 64, 171-192.

PERINCEK, D. & CEMEN, I. 1990. The structural relationship between the East Anatolian and Dead Sea fault zones in southeastern Turkey. Tectonophvsics, 172,331-340.

POLAT, A., KERRICH, R. & caseY, J.F. 1997. Geochemistry of Quaternary basalts erupted along the East Anatolian and Dead Sea fault zones of Southern Turkey: implications for mantle sources. Lilhos, 40, 55-68.

PONIKAROV, V., KOSLOV, V., ARTEMOV. A. & KALIS. A. 1963a. Geological Map of Syria, 1:200000, Horns (I-36-XVIII; I-37-XIII). Technoexport, Moscow.

PONIKAROV. V., SHATSKY, V., KAZMIN. V.. et al. 1963b. Geological Map of Syria, 1:200000. Hama (I-36-XX1V; 1-37-XIX). Technoexport, Moscow.

QUENNELL. A.M. 1983. Evolution of the Dead Sea Rift-a review. In: ABED, A.M. & KHALED, H.M. (eds) Geology of Jordan. Proceedings of the 1st Jordanian Geological Society Conference (Amman). Jordanian Geological Society, Amman, 460^t82.

QUENNELL, A.M. 1984. The Eastern Arabia rift system. In: DIXON, J.E. & ROBERTSON, A.H.F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 775-788.

RADWAN, Y., AL-NAIJAR. H. & LAYOUS, I. 1992. Indications of new tectonics along the Syrian Lebanese fault (El-Ghab depression). In: The Proceeding of the Regional Workshop on Archaeoseismicity in the Mediterranean Region, Damascus, 122-133.

ROJAY. B., HEIMANN. A. & TOPRAK. V. 2001. Neotectonic and volcanic characteristics of the Karasu fault zone (Anatolia, Turkey): the transition zone between the Dead Sea transform and the East Anatolian fault zone. Geodinamica Acta, 14, 197-212.

RON, H. 1987. Deformation along the Yammuneh. The restraining bend of the Dead Sea Transform: paleomagnetic data and kinematic implications. Tectonics, 6, 653-666.

RUKIEH, M. 1991. Report on the tectonism of the Western and Southern part of Syria and Lebanon by interpretation of space images. General Organisation of Remote Sensing, Damascus, 33.

RUKIEH. M., IBRAHIM, H. & HABIB, B. 1994. Remote sensing and study of volcanism in the southern part of the coastal range and its deeply xenolites (Syria). In: second International Conference on Geology of the Arab World. Cairo, 22-26 January 1994. General Organisation of Remote Sensing, Damascus. 1-11.

SALAMON, A., HOFSTETTER, A., GARFUNKEL, Z. & RON. H. 1996. Seismicity of the eastern Mediterranean region: perspective from the Sinai subplate. Tectonophysics, 263, 293-305.

SHARKOV, E.V., CHERNYSHEV, I.V. & DEVYATKIN. E.V. HT AL. 1994. Geochronology of Late Cenozoic basalts in Western Syria. Petrology, 2(4), 439^48.

SHARKOV, E.V., CHERNYSHEV, I.V. & DEVYATKIN, E.V. ET AL. 1998. New data on the geochronology of Upper Cenozoic plateau basalts from the northeastern periphery of the Red Sea rift area (Northern Syria). Doklady Earth Sciences, 358(l),19-22.

STEIN, M. & HOFMANN, A. 1992. Fossil plume head beneath the Arabian lithosphere? Earth and Planetary Science Letters, 114, 193-209.

WALLEY, C.D. 1998. Some outstanding issues in the geology of Lebanon and their importance in the tectonic evolution of the Levantine region. Tectonophvsics, 298, 37-62.

WALLEY, C.D. 2001. The Lebanon passive margin and the evolution of the Levantine Neo-Tethys. In: ZIEULER, P.A., CAVAZZA, W., ROBERTSON, A.H.F. & CRASQUIN-SOLF.AU, S. (eds) Peri-Tethys Memoir 6: Peri-Tethyan Rift/ Wrench Basins and Passive Margins. Mémoires du Muséum National d'Histoire Naturelle, 186, 407-439.

YURTMEN, S.. GUILLOU, H., WESTAWAY, R., ROWBOTHAM, G. & TATAR, O. 2002. Rate of strike-slip motion on the Amanos Fault (Karasu Valley, southern Turkey) constrained by K-Ar dating and geochemical analysis of Quaternary basalts. Tectonophysics, 344, 207-246.

YÜRÜR, T.M. & CHOROWICZ. J. 1998. Recent volcanism, tectonics and plate kinematics near the junction of the African, Arabian and Anatolian plates in the Eastern Mediterranean. Journal of Volcanology ami Geothermal Research, 85, 1-15.

ZAK. I. & FREUND, R. 1966. Recent strike-slip movement along the Dead Sea rift. Israel Journal of Earth Sciences, 15, 33-37.

ZILBERMAN. E., AMIT, R., HEIMANN. A. & PORAT, N. 2000. Changes in Holocene paleoseismic activity in the HuIa pull-apart basin. Dead Sea Rift, northern Israel. Teclonophysics, 312, 237-252.

Received 31 March 2004; revised typescript accepted 20 July 2004.

Scientific editing by Alex Mailman

J. CHOROWICZ1, D. DHONT2, O. AMMAR3, M. RUKIEH3 & A. BILAL4

1 UMR 7072: Laboratoire de Tectonique, case 129, Université Paris 6, 4 place Jussieu, 75252 Paris cedex 05, France (e-mail: jean.chorowicz@lgs.jussieu.fr)

2 UMR 5831: Imagerie géophysique, CURS-IPRA, Université de Pau et des Pays de l'Adour, Avenue de l'Université, 64013 Pau cedex, France

3 General Organization of Remote Sensing, PO Box 12586 Damascus, Syria

4 University of Damascus, Damascus, Syria

Copyright Geological Society Publishing House Mar 2005
Provided by ProQuest Information and Learning Company. All rights Reserved