Geodetic constraints on present day motion of the Arabian Plate: Implications for Red Sea and Gulf of Aden rifting

Transcription

1 Click Here for Full Article Geodetic constraints on present day motion of the Arabian Plate: Implications for Red Sea and Gulf of Aden rifting Abdullah ArRajehi, 1 Simon McClusky, 2 Robert Reilinger, 2 Mohamed Daoud, 3 Abdulmutaleb Alchalbi, 3 Semih Ergintav, 4 Francisco Gomez, 5 Jamal Sholan, 6 Firyal Bou Rabee, 7 Ghebrebrhan Ogubazghi, 8 Biniam Haileab, 8 Shimelles Fisseha, 9 Laike Asfaw, 9 Salah Mahmoud, 10 Ali Rayan, 10 Rebecca Bendik, 11 and Lewis Kogan 11 Received 6 March 2009; revised 29 October 2009; accepted 31 December 2009; published 24 June King Abdulaziz City for Science and Technology, Riyadh, Saudi Arabia. 2 Department of Earth, Atmospheric, and Planetary Sciences, MIT, Cambridge, Massachusetts, USA. 3 Syrian National Earthquake Center, Damascus, Syria. 4 TUBITAK Marmara Research Center, Gebze, Turkey. 5 Department of Geological Sciences, University of Missouri, Columbia, Missouri, USA. 6 Yemen National Seismological Observatory Center, Dhamar, Yemen. 7 Department of Geology, Kuwait University, Kuwait City, Kuwait. 8 Department of Earth Sciences, University of Asmara, Asmara, Eritrea. 9 Geophysical Observatory, AAU, Addis Ababa, Ethiopia. 10 NRIAG, Cairo, Egypt. 11 Department of Geosciences, University of Montana, Missoula, Montana, USA. Copyright 2010 by the American Geophysical Union /10/2009TC TECTONICS, VOL. 29,, doi: /2009tc002482, 2010 [1] Five years of continuously recording GPS observations in the Kingdom of Saudi Arabia together with new continuous and survey mode GPS observations broadly distributed across the Arabian Peninsula provide the basis for substantially improved estimates of present day motion and internal deformation of the Arabian plate. We derive the following relative, geodetic Euler vectors (latitude ( N), longitude ( E), rate ( /Myr, counterclockwise)) for Arabia Nubia (31.7 ± 0.2, 24.6 ± 0.3, 0.37 ± 0.01), Arabia Somalia (22.0 ± 0.5, 26.2 ± 0.5, 0.40 ± 0.01), Arabia India (18.0 ± 3.8, 87.6 ± 3.3, 0.07 ± 0.01), Arabia Sinai (35.7 ± 0.8, 17.1 ± 5.0, 0.15 ± 0.04), and Arabia Eurasia (27.5 ± 0.1, 17.6 ± 0.3, ± 0.004). We use these Euler vectors to estimate present day stability of the Arabian plate, the rate and direction of extension across the Red Sea and Gulf of Aden, and slip rates along the southern Dead Sea fault south of the Lebanon restraining bend ( ± 0.2 mm/yr, left lateral; ± 0.3 mm/yr extension) and the Owens fracture zone ( ± 0.5 mm/yr, right lateral, increasing from north to south; 1 2 mm/yr extension). On a broad scale, the Arabian plate has no resolvable internal deformation (weighted root mean square of residual motions for Arabia equals 0.6 mm/yr), although there is marginally significant evidence for N S shortening in the Palmyride Mountains, Syria at 1.5 mm/yr. We show that presentday Arabia plate motion with respect to Eurasia is consistent within uncertainties (i.e., ±10%) with plate tectonic estimates since the early Miocene when Arabia separated from Nubia. We estimate the time of Red Sea and Gulf of Aden rifting from present day Arabia motion, plate tectonic evidence for a 70% increase in Arabia Nubia relative motion at 13 Ma, and the width of the Red Sea and Gulf of Aden and find that rifting initiated roughly simultaneously (±2.2 Myr) along the strike of the Red Sea from the Gulf of Suez to the Afar Triple Junction, as well as along the West Gulf of Aden at 24 ± 2.2 Ma. Based on the present kinematics, we hypothesize that the negative buoyancy of the subducted ocean lithosphere beneath the Makran and the Zagros fold thrust belt is the principle driver of Arabia Eurasia convergence and that resisting forces associated with Arabia Eurasia continental collision have had little impact on plate motion. Citation: ArRajehi, A., et al. (2010), Geodetic constraints on present day motion of the Arabian Plate: Implications for Red Sea and Gulf of Aden rifting, Tectonics, 29,, doi: /2009tc Introduction [2] During the early Miocene ( 25 Ma), the Arabian plate separated from Africa along the Red Sea and Gulf of Aden rifts, roughly Ma prior to the initiation of Arabia Eurasia continental collision [e.g., McKenzie et al., 1970; Hempton, 1987; Joffe and Garfunkel, 1987; McQuarrie et al., 2003; Garfunkel and Beyth, 2006]. Separation from Africa resulted in the formation of ocean spreading in the Red Sea and Gulf of Aden [Chu and Gordon, 1998; Cochran, 1981] and in km of compression along a continental collision zone in eastern Turkey, the Zagros, and the Caucasus Mountains [e.g., Jolivet and Faccenna, 2000; McQuarrie et al., 2003]. This continental collision is a major driver of the active tectonics of the eastern Mediterranean region [e.g., Sengor et al., 1985, 2003; Allen et al., 2004] and the devastating earthquakes that have affected this area throughout recorded history [e.g., Ambraseys and Jackson, 1998]. Furthermore, the kinematics of the separation of Arabia from Africa, and the continuing continental collision with Eurasia, offer opportunities to evaluate the role of different forces in driving/resisting Arabia plate motion [McQuarrie et al., 1of10

2 Figure 1. GPS velocities (spatially decimated for clarity) in and adjacent to the Arabian plate with 1 sigma confidence ellipses in a Eurasia fixed reference frame. Velocities are tabulated in Data Set S1. Topography and bathymetry from Abbreviations are as follows: EAF, East Anatolian fault; GoS, Gulf of Suez; ATJ, Afar Triple Junction; ZFTB, Zagros foldthrust belt; GoA, Gulf of Aden; OFZ, Owens fracture zone. 2003; Bellahsen et al., 2003]. Precise constraints on Arabia plate motion are therefore important for evaluating earthquake hazards along Arabia s plate boundaries (Red Sea/ Gulf of Aden rifts, Dead Sea fault, East Anatolian fault, Zagros fold thrust belt, Makran subduction, and Owens fracture zone), and for constraining the dynamics of plate motions. [3] In this paper, we present new GPS constraints on Arabia plate motion, updating previously published results. We evaluate the impact of these new constraints on fault slip rates around the periphery of the plate, focusing primarily on rifting along the Red Sea and Gulf of Aden, and strike slip motion along the southern Dead Sea fault and the Owens fracture zone (Arabia India boundary). We compare presentday plate motions and fault slip rates to plate tectonic and geologic estimates and show that present day motions of the Arabian plate with respect to Eurasia are consistent (±10%) with these independent estimates since separation of Arabia from Africa in the early Miocene ( 25 Ma). Similarly, present day Nubian plate motion with respect to Eurasia is equal within uncertainties to plate motions estimated for at least the past 11 Ma. Finally, we speculate on the implications of the new GPS constraints for the dynamics of Arabia plate motion and collision with Eurasia. 2. GPS Data Analysis and Euler Vector Determination [4] Figure 1 shows GPS velocities used in this study (decimated spatially for clarity) in and adjacent to the Arabian plate along with 1 sigma confidence ellipses in a Eurasiafixed reference frame (the full velocity field is tabulated in Data Set S1). 1 We analyze the GPS data using the GAMIT/ GLOBK software [Herring, 2004; King and Bock, 2004] in a two step approach. The GPS solution is realized in the ITRF2005 global reference frame, and rotated into Eurasia, Nubia, Sinai, Somalia, India, and Arabia reference frames. Details of the processing strategy, error estimation, and reference frame definitions are identical to those used by Reilinger et al. [2006, Table S1] with updated velocity 1 Auxiliary materials are available at ftp://ftp.agu.org/apend/tc/ 2009tc of10

3 Figure 2. A simple block model for Arabian plate motion with residual velocities (i.e., component of Arabia site velocities not consistent with coherent plate motion) and 95% confidence ellipses. Velocity vectors with purple error ellipses were used to estimate the Arabian Plate reference frame. Red vectors indicate Arabian plate velocity at selected locations on the plate boundary. Present day rates of extension (mm/yr) across the Red Sea and Gulf of Aden, and slip rates (mm/yr) (P, fault parallel; N, fault normal) on the Dead Sea fault and Owens fracture zone derived from the block model (relative Euler vectors; Table 1) are also shown. The block model is derived from sites that are generally more than 50 km from plate boundaries (shown with purple error ellipses for Arabia and identified in Data Set S1 for other plates) and does not include estimates of strain accumulation (see text for discussion). The small residuals indicate that, except for elastic strains near plate boundaries and possible shortening across the Palmyride fold thrust belt [Alchalbi et al., 2010], the Arabian plate is not deforming internally at the present level of GPS uncertainties ( 1 mm/yr). Format as in Figure 1. estimates and are given in the online supplementary material for that paper (ftp://ftp.agu.org/apend/jb/2005jb004051). [5] To estimate relative Euler vectors, we use only GPS velocities at sites located more than 50 km from known plate boundaries (see Figure 2 for sites on Arabia and Data Set S1 for other plates) and search for the Euler vector that minimizes, in a least squares sense, the observed velocities within each plate interior (ts (this study) in Table 1). These Euler vectors allow us to estimate the present day rate and direction of relative motion across the Red Sea (Arabia Nubia) and Gulf of Aden (Arabia Somalia) as well as the Dead Sea fault (Arabia Sinai) and Owens Fracture zone (Arabia India). For the Dead Sea fault and Owens fracture zone, we decompose the relative motion at the location of the plate boundary faults into fault parallel and fault normal components to estimate strike slip and normal (i.e., extension or shortening) motion on the faults. We have chosen to use this approach, rather than an elastic block model [e.g., Meade and Hager, 2005; Reilinger et al., 2006] because of the concentration of GPS velocities along plate boundary faults (Figure 1). Any errors in defining the boundary fault (location, locking depth, dip) would result in mismodeling the elastic strain and could introduce spurious velocities for estimating plate motion. This approach works well for the larger plates (WRMS in Table 1), but is less reliable for the Sinai block (some sites used to define Sinai motion are less than 50 km from block boundaries) because elastic boundary deformation may affect the entire block. [6] Figure 2 shows residual velocities within Arabia and deduced boundary fault slip rates using the GPS Euler vectors 3of10

4 Table 1. Euler Vectors for Arabia With Respect to Neighboring Plates and 1 Sigma Uncertainties From This Study and Selected Other References a Plate Pair Longitude ( E) Latitude ( N) Rate ( CCW/Ma) WRMS of Plate 2 b (mm/yr) Number of Sites on Plate 2 b References c AR EU d 17.6 ± ± ± ts AR EU d 18.4 ± ± ± r AR EU d 18.4 ± ± ± 0.04 mc1 AR EU d ± ± ± se AR EU d 19.5 ± ± ± 0.1 ve AR EU 13.7 ± ± ± 0.05 nu AR NU d 24.6 ± ± ± ts AR NU d 25.2 ± ± ± r AR NU d 25.7 ± ± ± 0.04 mc1 AR NU d ± ± ± vi AR NU d ± ± ± se AR NU 23.0 ± ± ± 0.05 chu AR NU jg AR NU je AR SOM d 26.2 ± ± ± ts AR SOM d 27.4 ± ± ± r AR SOM d ± ± ± v AR SOM jg AR SOM ± ± ± 0.05 je AR SOM gd AR SOM gb AR SOM ± ± nu AR SIN d 17.1 ± ± ± ts AR SIN d 28.4 ± ± ± r AR SIN d ± ± ± wd AR SIN jg AR IN d 87.6 ± ± ± ts AR IN d ± ± ± 0.01 r AR IN d ± ± ± se AR IN d ± ± ± vi AR IN 92 ± ± ± 0.04 gd AR IN ± ± ± 0.04 nu AR IN f a CCW, counterclockwise. Key is as follows: EU, Eurasia; NU, Nubia; AR, Arabia; SIN, Sinai; SOM, Somalia; IN, India. b Twenty one sites on Arabia. WRMS is for sites used in the plate stabilization; WRMS for Arabia equals 0.5 mm/yr. Plate 2 is the neighboring plate with respect to Arabia. c References are as follows: ts, this study; chu, Chu and Gordon [1998]; f, Fournier et al. [2008]; gb, Garfunkel and Beyth [2006]; gd, Gordon and DeMets [1989]; je, Jestin et al. [1994]; jg, Joffe and Garfunkel [1987]; mc, McClusky et al. [2000]; mc1, McClusky et al. [2003]; nu, NUVEL1 A[DeMets et al., 1994]; r, Reilinger et al. [2006]; se, Sella et al. [2002]; ve, Vernant et al. [2004]; vi, Vigny et al. [2006]; wd, Wdowinski et al. [2004]. d Geodetic estimate. we estimate in this study (Table 1). Small, but significant residual velocities within the Arabian plate near the EAF are consistent with deformation due to elastic strain accumulation [Vigny et al., 2006; Reilinger et al., 2006]. Our estimated slip rates for the DSF south of the Lebanon restraining bend ( ± 0.2 mm/yr, left lateral; ± 0.3 mm/yr extension) are consistent with other geodetic estimates [e.g., Wdowinski et al., 2004; Mahmoud et al., 2005], although more tightly constrained, as well as with Late Pleistocene, geologic estimates [e.g., Klinger et al., 2000]. Similarly, our estimated slip rate for the Owens fracture zone ( ± 0.5 mm/yr, right lateral, increasing from north to south; 1 2 mm/yr extension) agrees with some prior geodetic and geologic estimates, as well as with the sense of motion deduced from earthquake focal mechanisms [DeMets et al., 1994; Reilinger et al., 2006; Fournier et al., 2008]. We do not report GPS slip rates for the central and northern Dead Sea fault [see Gomez et al., 2007; Alchalbi et al., 2010], the East Anatolian fault [Reilinger et al., 2006], or the Zagros fold thrust belt [Vernant et al., 2004; Walpersdorf et al., 2006; Tavakoli et al., 2008] since these slip rates depend on the detailed deformation of the Levant, Anatolian and Iran regions, respectively, that are not considered in this study. 3. Internal Deformation of the Arabian Plate [7] Figure 2 indicates that internal deformation of Arabia is small and mostly below the resolution of our GPS observations (i.e., 1 mm/yr). This includes sites broadly distributed around the plate from southeast Turkey (North Arabia), to central Arabia, to Oman (southeast Arabia), to Yemen (southwest Arabia). A single Euler vector accounts for almost all observed station motions (weighted root mean square (WRMS) of residual velocities for Arabian sites away from plate boundaries = 0.6 mm/yr). Exceptions to this are the three survey sites south of the Palmyride Mountains in Syria that show marginally significant northward motion relative to sites north of the Palmyride Mountains (Figure 2) [see also Alchalbi et al., 2010]. While we are reluctant to put too much confidence in this preliminary result, it finds support from palinspastic reconstructions of the Palmyride fold belt that estimate 20 km of shortening since early Miocene [Chaimov 4of10

5 Figure 3. Comparison of GPS fault slip rates derived from our block model (Figure 2 and Table 1) with geologic fault slip rates reported in the literature. The shaded area indicates ±10% from agreement. The geodetic rates for the EAF and the Gulf of Suez are from Reilinger et al. [2006], and that for the Zagros foldthrust belt is from Tavakoli et al. [2008]. Geologic references are as follows: Red Sea, Chu and Gordon [1998] and Joffe and Garfunkel [1987]; Gulf of Aden, Girdler et al. [1980], Laughton et al. [1970], and Cochran [1981]; Southern Dead Sea fault, Klinger et al. [2000] and Niemi et al. [2001]; Owens Fracture Zone, DeMets et al. [1994] and Fournier et al. [2008]; Main Recent fault, Authemayou et al. [2006]. et al., 1990], suggesting shortening rates of 1 mm/yr. However, differential motions across the Palmyrides indicated by our GPS observations are either too small to be determined on a plate wide basis or are entirely internal to the Arabian plate; that is, they do not change the overall plate configuration. If internal to the Arabian plate, any associated N S shortening in the Palmyride Mountains is presumably balanced by N S extension further south. 4. Geodetic Versus Geologic Relative Plate Motions and Fault Slip Rates [8] Plate tectonic estimates of Arabia plate motion have been reported from magnetic anomalies in the Red Sea covering the past 3 Ma [Chu and Gordon, 1998] and from broadscale plate reconstructions for the past 59 Ma [McQuarrie et al., 2003]. The 3 Ma Euler vector (Arabia Nubia) reported by Chu and Gordon [1998] differs insignificantly (1 sigma) in location and rotation rate from the GPS Euler vector reported here (Table 1) [see also McClusky et al., 2003]. Figure 3 shows a comparison of GPS fault slip rates derived from our block model (Figure 2) as well as from other geodetically constrained faults around the periphery of the Arabian plate, with recent geologic fault slip rates reported in the literature and covering a wide range of timescales ( years). Although reported, longer term fault slip rates vary considerably, for the most part they agree well with geodetic estimates (i.e., ±10%). A possible exception is the Main Recent Fault (Iran) that appears to have a higher geologic than geodetic slip rate. If this result persists with improved geologic observations, it could be indicative of complex fault interaction [Walpersdorf et al., 2006; Tavakoli et al., 2008]. [9] Figure 4 shows a plot of Africa/Arabia motion with respect to Eurasia as reported by McQuarrie et al. [2003] with the GPS plate rates for Africa (Nubia) and Arabia extrapolated to these geologic timescales. Present day Arabia plate motion (GPS) is equal within uncertainties to geologic estimates since at least 22 Ma, near the time when Arabia separated from Nubia. Similarly, present observations indicate 5of10

6 is about an order of magnitude less that the width of the Red Sea and Gulf of Aden basins. Accordingly, we estimate a 10% uncertainty on the degree to which our estimate of basin widths reflect total extension across these rifts. [11] Figure 6 shows a plot of the width of the Red Sea and Gulf of Aden basins estimated in Figures 5a and 5b versus the observed GPS velocity with respect to Nubia (Red Sea) and Somalia (Gulf of Aden). The linear relationship between velocity and basin width is well established. Following Le Pichon and Gaulier [1988], we estimate that the increase in Nubia motion with respect to Arabia initiated at 13 Ma. From Figure 4, we estimate a 70% decrease in Nubia Eurasia motion at that time (22.5 mm/yr to 6.6 mm/yr) while Arabia Eurasia motion remained approximately unchanged. If we assume that the change in Nubia Arabia motion was a change in rate only [McQuarrie et al., 2003, Figure S1], this implies a corresponding 70% increase in Arabia Nubia relative motion. Total extension of the Red Sea (Nubia Arabia boundary) will then be W ðtþ ¼ W ðpre13þ þ W ðpost13þ ð1þ Figure 4. Africa/Arabia motion [from McQuarrie et al., 2003] with the GPS plate rates for Africa (NU, red line) and Arabia (AR, blue line) extrapolated to these geologic timescales. The green line shows the rate of motion for Nubia at the prescribed location during the period from 30 to 13 Ma. The GPS plate rates are equal to within observational uncertainties to plate tectonic estimates of Africa motion for at least the past 11 Ma and for Arabia for more than 22 Ma, near the time when Arabia separated from Nubia. that Nubia motion has not changed significantly since at least 11 BP. 5. Implications for the Tectonic Evolution of the Red Sea and Gulf of Aden [10] The consistency of geodetic and geologic indicators of Arabia plate motion imply that highly precise, geodetically determined motions can be extrapolated to investigate the geologic evolution of plate boundary structures. Here we consider the Red Sea and Gulf of Aden rifts. Figure 5 shows GPS derived velocities for Arabia relative to Nubia along the Red Sea coast (Figure 5a) and for Arabia relative to Somalia along the Gulf of Aden coast (Figure 5b) and our estimate of the width of the rift basins in the direction of relative plate motion, and at the location of each GPS station along the Arabian coast. We estimate the width of the rift basins using the present day location of the coastline. This width is taken as a proxy for the total extension across the rift [e.g., McKenzie et al., 1970; Cochran, 1981; Sultan et al., 1993]. More precise estimates of total extension would require a detailed knowledge of the three dimensional geometry of all rift related faults and the total offset on each, information that is not presently available. However, we estimate that errors associated with our simple approach to determining basin width are likely to be less than the prerift crustal thickness that W ðtþ ¼ 0:3V ðgpsþ ðt 13Þ þ VðgpsÞ 13 where W (pre13) and W (post13) are the amount of opening of the Red Sea pre 13 Ma and post 13 Ma, W (t) is the total measured width of the Red Sea, V (gps) is the present day velocity (rate) of Arabia with respect to Nubia at the location of W (t), and T is the time of initial rifting. [12] Solving for T, T ¼ W ðtþ 9:1 0:3 ð3þ V ðgpsþ [13] From the slope of W (t) versus V (gps) in Figure 6 ( 16.4 ± 2.2 Myr), we estimate a time of initial opening of the Red Sea of 24 ± 2.2 Ma. While this age is uncertain because of uncertainty about the degree to which our estimates of W (t) reflect total extension, the slow rate of rifting prior to 13 Ma (and hence, T is highly sensitive to small changes in W (pre13) ), uncertainties about the exact timing of the change in Arabia Nubia relative motion, and our assumption that this change occurred instantaneously, it is roughly consistent with geologic estimates for the initiation of the main phase of rifting [e.g., Omar and Steckler, 1995;McQuarrie et al., 2003;Garfunkel and Beyth, 2006], demonstrating the internal consistency of our analysis. In addition, the uncertainty on the slope of W (t) versus V (gps) suggests that rifting initiated over a relatively short time (±2.2 Myr, or ±10% of the total time of rifting) along the full length of the Red Sea rift basin, a result that is supported by fission track analyses dating the early rift flank uplifts [Omar and Steckler, 1995]. 6. Discussion [14] The geodetic geologic comparisons presented in this paper strongly support the contention that the geodetic observations covering time spans of tens of years accurately ð2þ 6of10

7 Figure 5. (a) GPS velocities and 95% confidence ellipses with respect to Nubia along the west coast of the Arabian plate on a bathymetric and topographic map of the Red Sea and adjacent areas. Bold lines across the Red Sea rift are estimates of basin width at the location and direction of the GPS velocities. Basin widths (W (t) ) and observed Arabia plate velocities (V (gps) ) used to estimate the age and timing of extension in Figure 6 are also shown. Base map as in Figure 1. (b) As Figure 5a with GPS velocities with respect to Somalia. Base map as in Figure 1. reflect (i.e., ±10%) fault slip rates ( years) and geologic plate motions (10 7 years), and accordingly the longterm, geologic evolution of the Africa Arabia Eurasia zone of plate interaction [see also Allen et al., 2004; Reilinger et al., 2006]. Calais et al. [2003] suggest that Nubia Eurasia relative motion has slowed by about 1 mm/yr during the past 3 Ma based on comparison of updated NUVEL 1A, geologic estimates with geodetic estimates of Nubia Eurasia plate motion. However, owing to the present uncertainties in both the geodetic and geologic estimates used here, and the relatively slow, oblique convergence between Nubia and Eurasia (i.e., 4 5 mm/yr [McClusky et al., 2003]), a 1 mm/yr decrease in Nubia Eurasia relative motion would not be resolvable by our analysis. However, such a small change in plate rate, if real, would be within our estimated uncertainties and would not effect our conclusions. [15] As shown above, geologic and geodetic estimates of Arabia Eurasia relative plate motion agree to well below the observational uncertainties for the past 22 Ma and within 95% confidence uncertainties to 30 Ma. This timing corresponds 7of10

8 Figure 5. to the separation of Arabia from Africa and the initiation of rifting in the Red Sea and Gulf of Aden [e.g., Garfunkel and Beyth, 2006]. Furthermore, the present widths of the Red Sea and West Gulf of Aden are consistent with plate tectonic and geodetic estimates of Arabia Nubia motion over the past 24 ± 2.2 Ma. We find no evidence that Arabia plate motion with respect to Eurasia has slowed since separation from Nubia in the early Miocene. We suggest that the apparent discrepancies between the NUVEL 1A, 3 Ma estimate of Arabia Eurasia plate motion and geodetic estimates (i.e., Table 1) are most likely due to limitations of the geophysical observations that were used in the NUVEL model. [16] The steady motion of the Arabian plate since separation from Nubia and the initiation of continental collision in the Late Miocene/Pliocene, the decrease in the rate of convergence between Nubia and Eurasia during this same period [McQuarrie et al., 2003] (Figure 4), and the relatively short (continued) time estimated for the development of rifting along the Arabia Nubia and Arabia Somalia plate boundaries have direct implications for the dynamics of plate motions in the Arabia Africa Eurasia zone of plate interaction. As noted by McQuarrie et al. [2003], the slowing of Africa Eurasia convergence after separation of Arabia from Nubia in the Miocene is likely due to the loss of the subduction slab pull along the Arabia Eurasia plate boundary, strongly supporting dynamic models for plate motion where slab pull is the dominant plate driving force [e.g., Forsyth and Uyeda, 1975; Hager and O Connell, 1981; Conrad and Lithgow Bertelloni, 2004]. The steady motion of Arabia before, and after the initiation of continental collision with Eurasia seems to require that the resisting forces associated with continentcontinent collision are negligible in relation to the forces driving plate convergence, presumably the negative buoyancy of the subducted oceanic lithosphere. This may be due in 8of10

9 allowing rifting to concentrate along these structures [e.g., Garfunkel and Beyth, 2006]. Figure 6. Plot of the GPS Arabia plate motion rate with respect to Nubia (V (gps) ) versus the width of the Red Sea andgulfofadenrifts(w (t) ) at each GPS station location (from Figures 5a and 5b). The well defined, linear relationship between rift width and rate of extension is consistent with rifting initiating roughly simultaneously (±2.2 Myr, estimated from scatter around the straight line fit) along the Red Sea and the West Gulf of Aden at 24 ± 2.2 Ma (see text for discussion). part to weakening of the upper plate during the long history of Tethyan subduction and related back arc processes [Barazangi et al., 2006]. Furthermore, the absence of observable acceleration of Arabia following separation from Nubia suggests that the rate of plate motion depends primarily on the character of the subducting/subducted lithosphere and not on the dimensions of the trailing plate. Finally, the initiation of extension over a short time period along the strike of the Red Sea appears to be incompatible with dynamic upwelling associated with the Afar plume driving Arabia motion; such dynamics would more likely produce a tear that propagates from south to north [Burke and Dewey, 1973; Courtillot, 1982]. Again, we suggest that the negative buoyancy of the subducted ocean lithosphere beneath the Makran and Zagros are more likely responsible for the separation of Arabia from Nubia [Bellahsen et al., 2003]. On the other hand, the Afar plume may have weakened the lithosphere beneath the future Red Sea and Gulf of Aden rifts 7. Conclusions [17] In this study, we determine tightly constrained Euler vectors for Arabia plate motion relative to the Nubian, Sinai, Somalian, Eurasian, and Indian plates (Table 1). A single Euler vector describes well the motion of the Arabian plate indicating that any internal deformation of the plate is below the resolution of present geodetic data (i.e., < 1.5 mm/yr, or < 10% of Arabia plate motion rate relative to Eurasia). Based on these Euler vectors, we estimate present day slip rates for the Dead Sea fault south of the Lebanon restraining bend ( ± 0.2 mm/yr, left lateral; ± 0.3 mm/yr extension) and on the Owens fracture zone ( ± 0.5 mm/yr, right lateral, increasing from north to south; 1 2 mm/yr extension). Geodetically and geologically derived motions for Arabia and Nubia relative to Eurasia are equal within small uncertainties for at least the past 11 Ma for Nubia and > 22 Ma for Arabia, near the time when Arabia separated from Nubia in the early Miocene. We estimate the time rifting initiated in the Red Sea and the Gulf of Aden at 24 ± 2.2 Ma from the present width of the rifts, present day Arabia plate motion rates, and plate tectonic estimates of a 70% increase in Arabia Nubia relative motion at 13 Ma. Rifting appears to have initiated over a relatively short time (±2.2 Myr) along the full length of the Red Sea and along the West Gulf of Aden. We hypothesize that the kinematics of Arabia plate motion are most consistent with plate motion being driven by subduction processes (i.e., slab pull) along the Makran subduction zone and beneath the Zagros fold thrust belt, and that the Afar plume served mainly to weaken the African continental lithosphere allowing deformation to concentrate along the future Red Sea and Gulf of Aden rifts. [18] Acknowledgments. We thank those individuals and organizations that established and maintain the global GPS tracking network. We are grateful to Christophe Vigny, CNRS, France, for providing early GPS data from Yemen and to UNAVCO for assistance with the installation of the CGPS stations in Halyat Amar and Namas, KSA, and for logistical support for GPS survey observations. This research was supported in part by NSF grants EAR , EAR , and EAR and NASA grant NNG04GA15G to MIT, NSF grant EAR to the University of Missouri, and NSF grant EAR to the University of Montana. References Alchalbi, A., et al. (2010), Crustal deformation in northwestern Arabia from GPS measurements in Syria: Slow slip rate along the northern Dead Sea Fault, Geophys. J. Int., 180, , doi: / j x x. Allen, M., J. Jackson, and R. Walker (2004), Late Cenozoic reorganization of the Arabia Eurasia collision and the comparison of short term and longterm deformation rates, Tectonics, 23, TC2008, doi: /2003tc Ambraseys, N. N., and J. Jackson (1998), Faulting associated with historical and recent earthquakes in the eastern Mediterranean region, Geophys. J. Int., 133, , doi: /j x x. Authemayou, C., D. Chardon, O. Bellier, Z. Malekzadeh, E. Shabanian, and M. R. Abbassi (2006), Late Cenozoic partitioning of oblique plate convergence in the Zagros fold and thrust belt (Iran), Tectonics, 25, TC3002, doi: /2005tc Barazangi, M., E. Sandvol, and D. Seber (2006), Structure and tectonic evolution of the Anatolian Plateau in eastern Turkey, Spec. Pap. Geol. Soc. Am., 409, Bellahsen, N., C. Faccenna, F. Funiciello, J. M. Daniel, and L. Jolivet (2003), Why did Arabia separate from Africa? Insights from 3 D laboratory experiments, Earth Planet. Sci. Lett., 216, , doi: /s x(03) Burke, K., and J. F. Dewey (1973), Plume generated triple junctions: Key indicator in applying plate tectonics to old rocks, J. Geol., 81, , doi: / Calais, E., C. DeMets, and J. M. Nocquet (2003), Evidence for a post 3.16 Ma change in Nubia Eurasia North America plate motions?, Earth Planet. Sci. Lett., 216, 81 92, doi: /s x(03) Chaimov, T. A., M. Barazangi, D. Al Saad, T. Sawaf, anda.gebran(1990),crustalshorteninginthe Palmyride fold belt, Syria and implications for movement along the Dead Sea fault system, Tectonics, 9, Chu, D., and R. G. Gordon (1998), Current plate motions across the Red Sea, Geophys. J. Int., 135, , doi: /j x x. Cochran, J. R. (1981), The Gulf of Aden: Structure and evolution of a young ocean basin and continental margin, J. Geophys. Res., 86, , doi: /jb086ib01p Conrad, C. P., and C. Lithgow Bertelloni (2004), The temporal evolution of plate driving forces: Importance of slab suction versus slab pull during 9of10

10 the Cenozoic, J. Geophys. Res., 109, B10407, doi: /2004jb Courtillot, V. (1982), Propagating rifts and continental breakup, Tectonics, 1, , doi: / TC001i003p DeMets, C., R. G. Gordon, D. F. Argus, and S. Stein (1994), Effect of recent revisions to the geomagnetic reversal time scale on estimates of current plate motions, Geophys. Res. Lett., 21, , doi: /94gl Forsyth, D., and S. Uyeda (1975), On the relative importance of the driving forces of plate motion, Geophys. J. R. Astron. Soc., 43, Fournier, M., N. Chamot Rooke, C. Petit, O. Fabbri, P. Houchon, B. Maillot, and C. Lepvrier (2008), In situ evidence for dextral active motion at the Arabia India plate boundary, Nat. Geosci., 1, 54 58, doi: /ngeo Garfunkel, Z., and M. Beyth (2006), Constraints on the structural development of Afar imposed by the kinematics of the major surrounding plates, in The Afar Volcanic Province Within the East African Rift System, edited by G. Yirgu, C. J. Ebinger, and P. K. H. Maquire, Geol. Soc. Spec. Publ., 259, 23 42, doi: gsl.sp Girdler, R. W., C. Brown, D. J. M. Noy, and P. Styles (1980), A geophysical survey of the western Gulf of Aden, Philos. Trans. R. Soc. London, Ser. A, 298, 1 43, doi: /rsta Gomez, F., et al. (2007), Global Positioning System measurements of strain accumulation and slip transfer through the restraining bend along the Dead Sea fault system in Lebanon, Geophys. J. Int., 168, , doi: /j x x. Gordon, R. G., and C. DeMets (1989), Present day motion along the Owens Fracture Zone and Dalrymple Trough in the Arabian Sea, J. Geophys. Res., 94, , doi: /jb094ib05p Hager, B. H., and R. J. O Connell (1981), A simple global model of plate dynamics and mantle convection, J. Geophys. Res., 86, , doi: / JB086iB06p Hempton, M. R. (1987), Constraints on Arabian Plate motion and extensional history of the Red Sea, Tectonics, 6, , doi: /tc006i006p Herring, T. A. (2004), GLOBK: Global Kalman filter VLBI and GPS analysis program version 4.1, Mass. Inst. of Technol., Cambridge. Jestin, F., P. Huchon, and J. Gaulier (1994), The Somalia plate and the East Africa rift system: Present day kinematics, Geophys. J. Int., 116, , doi: /j x.1994.tb03286.x. Joffe, S., and Z. Garfunkel (1987), Plate kinematics of the circum Red Sea A re evaluation, Tectonophysics, 141, 5 22, doi: / (87) Jolivet, L., and C. Faccenna (2000), Mediterranean extension and the Africa Eurasia collision, Tectonics, 19, , doi: /2000tc King, R. W., and Y. Bock, (2004), Documentation of the MIT GPS analysis software: GAMIT, Mass. Inst. of Technol., Cambridge. Klinger, Y., J. P. Avouac, N. Abou Karaki, L. Dorbath, D. Bourles, and J. L. Reyes (2000), Slip rate on the Dead Sea transform fault in the northern Araba valley (Jordan), Geophys. J. Int., 142, , doi: /j x x. Laughton, A. S., R. B. Whitmarsh, and M. T. Jones (1970), The evolution of the Gulf of Aden, Philos. Trans. R. Soc. London, Ser. A, 267, , doi: /rsta Le Pichon, X., and J. M. Gaulier (1988), The rotation of Arabia and the Levant fault system, Tectonophysics, 153, , doi: / (88) Mahmoud, S., R. Reilinger, S. McClusky, P. Vernant, and A. Tealeb (2005), GPS evidence for northward motion of the Sinai block: Implications for E. Mediterranean tectonics, Earth Planet. Sci. Lett., 238, , doi: /j.epsl McClusky, S., et al. (2000), Global Positioning System constraints on plate kinematics and dynamics in the eastern Mediterranean and Caucasus, J. Geophys. Res., 105, McClusky, S., R. Reilinger, S. Mahmoud, D. Ben Sari, and A. Tealeb (2003), GPS constraints on Africa (Nubia) and Arabia plate motion, Geophys. J. Int., 155, , doi: /j x x. McKenzie,D.P.,D.Davies,andP.Molnar(1970), Plate tectonics of the Red Sea and east Africa, Nature, 226, , doi: /226243a0. McQuarrie, N., J. M. Stock, C. Verdel, and B. P. Wernicke (2003), Cenozoic evolution of Neotethys and implications for the causes of plate motions, Geophys. Res. Lett., 30(20), 2036, doi: / 2003GL Meade, B. J., and B. H. Hager (2005), Block models of crustal motion in southern California constrained by GPS measurements, J. Geophys. Res., 110, B03403, doi: /2004jb Niemi, T. M., H. Zang, M. Atallah, and J. B. J. Harrison (2001), Late Pleistocene and Holocene slip rate of the northern Wadi Araba fault, Dead Sea Transform, Jordan, J. Seismol., 5, , doi: / A: Omar,G.I.,andM.S.Steckler(1995),Fissiontrack evidence on the initial rifting of the Red Sea: Two pulses, no propagation, Science, 270, , doi: /science Reilinger, R., et al. (2006), GPS constraints on continental deformation in the Africa Arabia Eurasia continental collision zone and implications for the dynamics of plate interactions, J. Geophys. Res., 111, B05411, doi: /2005jb Sella, G. F., T. H. Dixon, and A. Mao (2002), REVEL: A model for recent plate velocities from space geodesy, J. Geophys. Res., 107(B4), 2081, doi: / 2000JB Sengor, A. M. C., N. Gorur, and F. Saroglu (1985), Strike slip faulting and related basin formation in zones of tectonic escape: Turkey as a case study, in Strike Slip Deformation, Basin Formation, and Sedimentation, edited by K. T. Biddle and N. Christie Blick, Spec. Publ. Soc. Econ. Paleontol. Mineral., 37, Sengor,A.M.C.,S.Özeren,T.Genç,andE.Zor (2003), East Anatolian high plateau as a mantlesupported, north south shortened domal structure, Geophys. Res. Lett., 30(24), 8045, doi: / 2003GL Sultan, M., R. Becker, R. E. Arvidson, P. Shore, R. J. Stern, Z. El Alfy, and R. I. Attia (1993), New constraints on Red Sea rifting from correlations of Arabian and Nubian Neoproterozoic outcrops, Tectonics, 12, ,doi: /93TC Tavakoli, F., A. Walpersdorf, C. Authemayou, H. R. Nankali, D. Hatzfeld, M. Tatar, Y. Djamour, F. Nilforoushan, and N. Cotte (2008), Distribution of the right lateral strike slip motion from the Main Recent Fault to the Kazerun fault system (Zagros, Iran): Evidence from present day GPS velocities, Earth Planet. Sci. Lett., 275, , doi: /j.epsl Vernant, P., et al. (2004), Present day crustal deformation and plate kinematics in the Middle East constrained by GPS measurements in Iran and northern Oman, Geophys. J. Int., 157, , doi: /j x x. Vigny, C., P. Huchon, J. C. Ruegg, K. Khanbari, and L. M. Asfaw (2006), Confirmation of Arabia plate slow motion by new GPS data in Yemen, J. Geophys. Res., 111, B02402, doi: /2004jb Walpersdorf, A., D. Hatzfeld, H. Nankali, F. Tavakoli, F. Nilforoushan, M. Tatar, P. Vernant, J. Chéry, and F. Masson (2006), Difference in the GPS deformation pattern of North and central Zagros (Iran), Geophys. J. Int., 167, , doi: / j x x. Wdowinski, S., Y. Bock, G. Baer, L. Prawirodirdjo, N. Bechor, S. Naaman, R. Knafo, F. Forrai, and Y. Melzer (2004), GPS measurements of current crustal movements along the Dead Sea fault, J. Geophys. Res., 109, B05403, doi: /2003jb A. Alchalbi and M. Daoud, Syrian National Earthquake Center, Damascus, Syria. (m daoud@scsnet.org) A. ArRajehi, King Abdulaziz City for Science and Technology, Riyadh 11442, Saudi Arabia. (arrajehi@ kacst.edu.sa) L. Asfaw and S. Fisseha, Geophysical Observatory, AAU, Addis Ababa, Ethiopia. (wmikel99@yahoo.com) R. Bendik and L. Kogan, Department of Geosciences, University of Montana, Missoula, MT 59812, USA. (bendick@mso.umt.edu) F. Bou Rabee, Department of Geology, Kuwait University, Kuwait City 13060, Kuwait. (firyal@ gmail.com) S. Ergintav, TUBITAK Marmara Research Center, Gebze, Turkey. (semih.ergintav@mam.gov.tr) F. Gomez, Department of Geological Sciences, University of Missouri, Columbia, MO 65211, USA. (fgomez@missouri.edu) B. Haileab and G. Ogubazghi, Department of Earth Sciences, University of Asmara, Asmara, Eritrea. (ghebrebrhanog@yahoo.com) S. Mahmoud and A. Rayan, NRIAG, Helwan, Cairo 11722, Egypt. (salahm@nriag.sci.eg) S. McClusky and R. Reilinger, Department of Earth, Atmospheric, and Planetary Sciences, MIT, Cambridge, MA 02139, USA. (reilinge@erl.mit.edu) J. Sholan, Yemen National Seismological Observatory Center, Dhamar, Yemen. (sholan20@ hotmail.com) 10 of 10