T23D-01 INVITED
Seismic Tomography Of The Arabian-Eurasian Collision Zone And Tectonic Implications
The objective of our research is to determine P and S wave velocity structures in the crust and upper mantle in the Arabian-Eurasian collision zone and surrounding areas, including Iran, Arabia, Eastern Turkey, the Caucasus, and Central Asia. In recent years the number of seismic stations has increased greatly in the region because of new and expanded seismic networks in Azerbaijan, Turkey, Iran and other countries in the region. Consequently, a large number of new P and S wave phase data has become available for tomography. Travel-time tomography is carried out by obtaining first the crustal thickness and Pn and Sn velocities from local and regional arrival time data. Then the P-velocity model is extended into the upper mantle by combining local, regional, and teleseismic data and crust model constraints. Pn velocity tomograms were obtained using 160,000 arrival times from 850 stations and 18,000 earthquakes. For Sn tomography, 75,000 phase readings were used. The Pn and Sn velocity models agree quite well although there are some local differences. Pn velocities are very low under eastern Anatolia, northwest Iran, and the Lesser Caucasus. There are localized low velocity anomalies. Velocities are low under the Iranian plateau. Pn velocities are high under the Arabian Platform, the Gulf, and the Zagros. In the north, there is an east-west trending narrow zone of high Pn velocities that includes eastern Black Sea, Kura Basin between Greater and Lesser Caucasus, south Caspian sea, and Kara and Kizil Kum Basins in Central Asia. The upper mantle tomograms show the images of the subducted Neotethys slab. The slab geometry is quite complex, reflecting the history of the changes in the plate motions and collision processes.
T23D-02
Lg Attenuation Modeling in the Middle East
We present a broadband tomographic model of Lg attenuation in the Middle East derived from source- and site-corrected amplitudes. The study region spans from Turkey through the Arabian Peninsula and Iran to Pakistan, Afghanistan, and northwest India. Absolute amplitude measurements are made on hand-selected and carefully windowed seismograms for tens of stations and thousands of crustal earthquakes resulting in excellent coverage of the region. We have modified the standard attenuation tomography technique to more explicitly define the earthquake source expression in terms of the seismic moment. This facilitates the use of the model to predict the expected amplitudes of new events, an important consideration for earthquake hazard or explosion monitoring applications. We will discuss the updated method and implications of this parameterization. A conjugate gradient method is used to tomographically invert the amplitude dataset of over 8000 paths. We solve for Q variation, as well as site and source terms, for a wide range of frequencies ranging from 0.5 –- 10 Hz. The attenuation results have a strong correlation to tectonics. Shields have low attenuation, while tectonic regions have high attenuation, with the highest attenuation at 1 Hz found in eastern Turkey. The results also compare favorably to other studies in the region made using Lg propagation efficiency, Lg/Pg amplitude ratios and two-station methods. We tomographically invert the amplitude measurements for each frequency independently. In doing so, it appears the frequency-dependence of attenuation is not compatible with the power law representation of Q(f). This research was performed under the auspices of the U.S. Department of Energy by the Lawrence Livermore National Laboratory under contract number DE-AC52-07NA27344. This is LLNL contribution LLNL-ABS-406761.
T23D-03
Structure of the Lithosphere and Upper Mantle Across the Arabian Peninsula
Tectonic structure of the Arabian Peninsula is revealed by analysis of modern broadband (BB) waveform data from the region. This presentation will report inferences of seismic structure of the Arabian Peninsula using various techniques and interpretations of its tectonic significance. Several studies have been undertaken using data from the Saudi Arabian National Digital Seismic Network (SANDSN) which consists of 38 (26 BB, 11 SP) stations, mostly located on the Arabian Shield. Additional data were taken from the 1995- 7 Saudi Arabian IRIS-PASSCAL Deployment (9 BB stations) and other stations across the Peninsula. Crustal structure, inferred from teleseismic P-wave receiver functions, reveals thicker crust in the Arabian Platform (40-45 km) and the interior of the Arabian Shield (35-40 km) and thinner crust along the Red Sea coast. Lithospheric thickness inferred from teleseismic S-wave receiver functions reveals very thin lithosphere (40- 80 km) along the Red Sea coast, which thickens rapidly toward the interior of the Arabian Shield (100-120 km). We also observe a step of 20-40 km in lithospheric thickness across the Shield-Platform boundary. Seismic velocity structure of the deep upper mantle inferred from teleseismic P- and S-wave travel time tomography reveals large differences between the Shield and Platform, with the Shield being underlain by slower velocities, 3 and 6 percent for P- and S-waves, respectively. Shallower upper mantle structure inferred from Rayleigh wave phase velocities indicates that low-velocity material is channeled beneath the Red Sea below 150 km and pools beneath the Arabian Shield above this depth. Seismic anisotropy inferred from shear-wave splitting reveals a splitting time of approximately 1.4 seconds, with the fast axis slightly east of north. The shear-wave splitting results are surprisingly consistent across the Peninsula, with a slight clockwise rotation parallel to the Dead Sea Transform Fault for stations near the Gulf of Aqaba. These results allow us to make several conclusions about the tectonic evolution and current state of the Arabian Plate. Lithospheric thickness implies that thinning near the Red Sea has accompanied the rupturing of the Arabian-Nubian continental lithosphere. The step in the lithospheric thickness across the Shield- Platform boundary likely reveals a pre-existing difference in the lithospheric structure prior to accretion of the terranes composing the eastern Arabian Shield. Tomographic imaging of upper mantle velocities implies a single large-scale thermal anomaly underlies the Arabian Shield and is associated with Cenozoic uplift and volcanism. Hot material from the Afar Hot Spot is channeled along the Red Sea by lithospheric topography.
T23D-04 INVITED
Lithospheric Structure of the Zagros and Alborz Mountain Belts (Iran) from Seismic Imaging
We present a synthesis of the results of two dense temporary passive seismic experiments installed for a few months across Central Zagros for the first one, and from North-western Zagros to Alborz for the second one. On both transects, the receiver function analysis shows that the crust has an average thickness of ~ 43 km beneath the Zagros fold-and-thrust belt and the Iranian plateau. The crust is thicker in the back side of the Main Zagros Reverse Fault (MZRF), with a larger maximum Moho depth in Central Zagros (69 ± 2 km) than in North-western Zagros (56 ± 2 km). To reconcile Bouguer anomaly data and Moho depth profile of Central Zagros, we proposed that the thickening is related to overthrusting of the Arabian margin by Central Iran on the MZRF considered as a major thrust fault rooted at Moho depth. The better-quality receiver functions of NW Zagros display clear conversions on a low-velocity channel which cross-cuts the whole crust from the surface trace of the MZRF to the Moho on 250-km length. Waveform modeling shows that the crustal LVZ is ~ 10-km thick with a S-wave velocity 8-30 % smaller than the average crustal velocity. We interpret the low-velocity channel as the trace of the thrust fault and the suture between the Arabian and the Iranian lithospheres. We favour the hypothesis of the LVZ being due to sediments of the Arabian margin dragged to depth during the subduction of the Neotethyan Ocean. At upper mantle depth, we find shield-like shear-wave velocities in the Arabian upper-mantle, and lower velocities in the Iranian shallow mantle (50-150 km) which are likely due to higher temperature. The lack of a high-velocity anomaly in the mantle northeast of the MZRF suture suggests that the Neotethian oceanic lithosphere is now detached from the Arabian margin. The crust of the Alborz mountain range is not thickened in relation with its high elevations, but its upper mantle has low P-wave velocities.
T23D-05
HIGH-RESOLUTION SEISMIC VELOCITY MODEL OF THE CAUCASUS-CASPIAN REGION
The Caucasus-Caspian region is part of the Alpine-Himalayan collision belt and is an area of complex structure accompanied by large variations in seismic wave velocities. Despite the complexity of the region little is known about the lithospheric structure. Using data from 25 new broadband seismic stations in the region, a unified velocity structure is developed using teleseismic receiver functions and surface waves. Several distinct regions are recognized: the Greater Caucasus, the lesser Caucasus and the Caspian/Kura Depression. Depth(h)-Vp/Vs(k) stacks of the receiver functions suggest crustal thicknesses of 45-50 km in the Lesser and Greater Caucasus with relatively high crustal velocities (Vs=3.4 km/s). Variations of the receiver functions with back-azimuth indicate 3D variations in structure in the Greater Caucasus. Crustal thickness in the Kura Depression and at the edge of the Caspian is less well constrained due to pronounced multiples associated with thick sedimentary layers but the preferred results suggest a Moho depth of 38 to 48 km and lower crustal velocities (Vs=3.1 km/s). Love and Rayleigh surface waves dispersion curves have been derived from using both event-based analysis and ambient noise correlation. Short-period surface waves outline the sedimentary structure of the region, while longer periods highlight crustal velocity, crustal thickness, and upper mantle velocity. Joint inversion of the receiver functions with the surface wave dispersion curves to determine absolute shear wave velocity is ongoing.
T23D-06 INVITED
Seismotectonics of the Iran Region
More than 2000 instrumentally recorded earthquakes occurring in the Iran region during the period 1918- 2008 have been relocated using an advanced seismic location technique. Relocation sharpens the image of seismic activity in the region and – more importantly – significantly improves event focal depths. Iranian seismicity is largely a result of the early stages of continent/continent collision (25-35mm/yr of northwards overall shortening) between the Arabian Peninsula and Eurasia. Most earthquakes in the Iranian continental lithosphere occur in the upper crust (consistent with focal depths of available local seismic network hypocenters), with crustal shortening accommodated entirely by thickening and distributed deformation. This shortening across Iran results in thrust and strike-slip faulting. In the Zagros Mountains nearly all earthquakes are confined to the upper crust (depths < 20 km), and there is no evidence for a seismically active subducted slab dipping NE beneath central Iran. Moreover, the Zagros has many earthquakes but their magnitudes are all less than Mw 7.0 and nearly all the moment release occurs near the SW topographic edge (i.e., elevations between 500-1000m) of the belt. The moment release in the Zagros cannot account for the expected convergence across it, suggesting that the missing moment release is being accommodated aseismically. In southeastern Iran, where the Arabian seafloor is being subducted beneath the Makran coast, low-level earthquake activity occurs in the upper crust as well as to depths of at least 150 km within a northward-dipping subducting slab. Near the Oman Line, a region transitional between the Zagros and the Makran, seismicity extends to depths of up to 30-45 km in the crust, consistent with low-angle thrusting of Arabian basement beneath central Iran. In north-central Iran, along the Alborz mountain belt, seismic activity occurs primarily in the upper crust but with some infrequent events in the lower crust, particularly in the western part of the belt (the Talesh), where the South Caspian basin underthrusts NW Iran. Earthquakes that occur in a band across the central Caspian, following the Apscheron-Balkhan sill between Azerbaijan and Turkmenistan, have depths in the range 30-100 km, deepening northwards. These are thought to be connected with either incipient or remnant northeast subduction of the South Caspian basin basement beneath the east-west trending Apscheron-Balkhan sill. Curiously, in this region of genuine mantle seismicity, there is no evidence for earthquakes shallower than 30 km.
T23D-07
Thick lithospheric root beneath the Zagros Mountains of Iran
The Zagros form one of the largest and youngest collisional orogenic belts on Earth but knowledge of how the shortening is accommodated across the Zagros is poorly understood. At shallow depths the shortening is accommodated by folding in the sediments and by high-angle reverse faulting in the basement. How the shortening is accommodated in the uppermost mantle has been uncertain largely because the seismic structure of the upper mantle beneath the region was poorly known. We map the lateral variation in the upper mantle lid beneath this region using a surface wave data set consisting of ~20,000 vertical component multi-mode waveforms. At shallow depths (75-100 km) the Iranian Plateau is slow. At deeper depths (150-225 km) the upper mantle is slow beneath Arabia and NE Iran but is fast below the Mesopotamian foredeep and the Zagros Mountains. This implies that the shortening in the mantle across the Zagros is accommodated by the formation of a thick lithospheric root beneath the Zagros either by underthrusting Arabian continental lithosphere beneath Iran, or by homogeneous deformation of the lithosphere. The extent of this thick lithospheric root beneath the Zagros correlates with variations in the geologic and tectonic process occurring in the Iranian Plateau.
T23D-08
Towards Thermomechanical Model of the Entire Dead Sea Transform
Dead Sea Transform (DST) fault system is a part of the Syrian-African rift system and it extends from the divergent plate boundary of Red Sea rift at the south to the convergent plate boundary in the Taurus Mountains at the north. DST is a left-lateral transform fault, accommodating differential motion between African and Arabian plates. The Eulerian pole of the relative plate motion is defined at 32.8° N 22.6° E. The morphology of the DST fault system is expressed by several linear stretches separated by a number of pull-apart basins, where the Dead Sea is a largest. Our previous models (Sobolev et al. 2005, Petrunin and Sobolev, 2006,2008) have been focused at two main topics: (1) major controls of the fault localization in strike slip settings and (2) major controls of the structure and evolution of pull-apart basins located at strike- slip faults. To do so, we use realistic elasto-visco-plastic temperature and stress dependent rheology to model lithospheric deformations. The largest limitation of our models appears to be their relatively small size, which does not allow including the source of the strike-slip motion in the region, which is likely opening of the Red Sea Rift, and major obstacle for the propagating fault resulting in its bending in Lebanon. In present work we extend the model to the larger region. The new model domain includes north-western part of the Red Sea and extends to the Lebanon Mountains in the north where deformation becomes more complicated and large part of the strike-slip motion becomes distributed. Because of the significant size of the domain, we made an improvement in the modelling technique taking into account sphericity of the Earth surface. Here we will show first results of our modelling aimed at revealing controls of localisation of the DST and origination of the chain of pull-apart basins in its southern part and transpression in the Lebanon.