GEOLOGICAL SETTING

The North Anatolian Fault (NAF) is a major dextral transform fault controlling the westward motion of the Anatolian Plate [Armijo et al.(1999)]. It extends for more than 1.600 km from the Karliova triple junction in the east, to the northern Aegean Sea in the west [Sengor (1979),Barka (1992),Cagatay et al.(1998),Hubert-Ferrari et al.(2002),Sengor et al.(2004)]. It is a widely held notion that the NAF originated in the Miocene following the development of the Bitlis suture along the Arabia-Eurasia collision zone (e.g. [Sengor et al.(1985)]) and reached the Marmara-Aegean region during the Pliocene (5 Ma; [Armijo et al.(1999)] or 3.4 Ma; [Yaltirak et al.(2000)]).

East of the Marmara Sea the NAF is constituted by a single strand characterized by a narrow deformation zone [Hubert-Ferrari et al.(2002)]. In the Sea of Marmara region the right lateral NAF splays into two major fault branches about 100 km apart. According to geological and geodetic data [Armijo et al.(1999),McClusky et al.(2000)] most of the lateral motion is transferred obliquely from the southern to the northern branch, across the Marmara basin [Armijo et al(2002)].

Different structural models have been put forward for the NAF in the Marmara Sea region and the tectonic regime is described as :

1. resulting from the interaction of Anatolia strike slip with Aegean extension [Wong et al.(1995),Parke et al.(1999)]
2. a single through-going master strike slip fault [Le Pichon et al.(2001),Le Pichon et al.(2003),Aksu et al.(2000),Imren et al.(2001),Gokasan et al.(2001),Kuscu et al.(2002),Alpar and Yaltirak (2002),Demirbag et al.(2003)]
3. a pull apart system produced by fault segmentation, oversteps and slip partitioning [Armijo et al.(1999),Armijo et al(2002),Barka and Kadisky-Cade (1988),Barka (1992),Stein et al.(1997),Okay et al.(2000),Parke et al.(2002),Flerit et al.(2003),Polonia et al.(2004)]

Geodetic measurements [McClusky et al.(2000)] suggest a 24 mm/yr right lateral motion between Anatolia and Eurasia in this region, with more than 80 % of the total motion, i.e., $>$ than 20 mm/y, along the northern branch [Armijo et al(2002),Meade et al.(2002)]. GPS geodetic measurements cover a period of ten years, a short interval of time compared to recurrence period of major earthquakes in the region [Ambraseys and Finkel (1991)]. Recently, [Povost et.(2003)] through a 3-D mechanical modeling of the GPS velocity field, provided an estimate of the NAF slip rate in the Marmara Sea in the order of 17.5 mm/yr, which is a value significantly smaller than the 24 mm/yr rate of [McClusky et al.(2000)].

Long term slip rate estimates from geological reconstructions [Armijo et al.(1999),Shindler (1997)] are systematically smaller than 24 mm/yr being in the range of 14-20 mm/yr during the last 3-4 Ma. The slip rate over a very large time scale (5 Ma) has been reconstructed through the analysis of displaced geological features in the Ganos region [Armijo et al.(1999)]. This estimate implies a total dextral slip of about 85 km in the past 5 Ma. However, the use of the particular displaced geological features has been questioned by [Okay et al.(2004)].

Estimates of the slip rate along the NAF over geological time scales and distribution of motion along the various strands of the NAF can help reach a realistic assessment of seismic hazards for this densely populated area of Turkey. During MARM2000 and MARM2001 cruises we obtained high-resolution acoustic images of the NAF in the floor of the eastern Marmara Sea (Gulf of Izmit), and measured fault-related offsets of $^{14}$C dated subseafloor channels and paleo-shorelines [Polonia et al.(2002)]. The resulting average slip rate on the fault is 10 mm/yr for the last 10 kyr (see Fig.4 and [Polonia et al.(2004)]). This is less than half the total Anatolia-Eurasia relative motion, estimated from satellite geodetic measurements.

figure

Figure 4: CHIRP SBP data analysis (MARM2001): fault-related offsets of $^{14}$C dated subseafloor channels and paleo-shoreline [Polonia et al.(2002),Polonia et al.(2004),Cagatay et al.(2000),Cagatay et al.(2003)]

The lack of understanding of the structural configuration and deformation rates of the NAF in the Marmara Sea has important effects on seismic hazard estimates for this densely populated region. Key areas for determing whether or not the NAF is segmented and at which scale such segmentation occurs are the Bay of Izmit  in the east and the Ganos Basin and Saros Gulf, in the west, as well as the surrounding regions on land. These two areas characterize the transition from an almost purely transcurrent regime to transtensional tectonics.

The Urania 2005 cruise has been planned in order to acquire new geological-geophysical data in these key areas. The new dataset will be integrated with preexisting data gathered by us and by other researchers, thus connecting detailed neotectonic, paleoseismological, and seismological information on Quaternary events with low-temperature thermochronology and seismic structural-stratigraphic interpretations on a longer time range, to ultimately elucidate the style of the transition between the strike-slip- dominated and the extensional regimes and to constrain the development of the NAF system.

Morphobathymetric and seismic surveys made by French oceanographic vessels [Le Pichon et al.(2005),,Armijo et al.(2005)] imaged in detail the surficial and deep structure of the sedimentary basins located in the deepest portions of the Marmara Sea. Despite these surveys, there are still unresolved disputes concerning the geological evolution of the deep basins. During cruises MARM2000 and MARM2001 the seismogenic branches of the NAF in shallow waters were identified and their kinematics was quantified. This project aims at integrating data at various spatial and temporal scales. The study of the geometry and evolution of transform faults requires the correlation of surficial and deep structures and the reconstruction of the temporal evolution of the deformation across the entire section of the fault system. High-resolution analysis of seismic images is necessary in order to reconstruct the deformational regime on the fault plane (purely transcurrent, transtensional, or transpressional) as determined by fault geometry and the stress field. This is the only approach to reconstruct the degree of deformation, the strain rates and their variations through time, the strain transfer on adjacent fault planes, and the formation of the various fault systems, thus identifying the segments most at risk.

The Marmara region is characterized by intense crustal fragmentation, with discrete blocks, tectonic rotations, and complex transtensional and transpressional movements. Integrated geophysical studies at various scales of resolution are not presently available. In order to characterize the complexity of the Marmara region we will compare geological/geophysical data at various scales and correlate the structures on land and at sea. For this reason, we propose to integrate the available marine seismic profiles with high-resolution (from a few decimeters to a few meters) geophysical investigations both on land and at sea. These geophysical data have the potential to resolve the geometry of single fault branches of the NAF system and will be integrated by geological data on-shore (field mapping, structural analysis of brittle deformation, low-temperature thermochronology) and off-shore (deep coring down to 20 m, well-logs data).