Constraints on North Anatolian Fault Zone Width in the Crust and Upper Mantle From S Wave Teleseismic Tomography

Abstract

We present high‐resolution S wave teleseismic tomography images of the western segment of the North Anatolian Fault Zone (NAFZ) in Turkey using teleseismic data recorded during the deployment period of the Dense Array for Northern Anatolia array. The array comprised 66 stations with a nominal station spacing of 7 km, thus permitting a horizontal and vertical resolution of approximately 15 km. We use the current S wave results with previously published P wave teleseismic tomography to produce maps of relative VP/VS anomalies, which we use to highlight the difference in overall composition of the three terranes separated by the northern (NNAF) and southern branches of the NAFZ. Our results show a narrow S wave low‐velocity anomaly beneath the northern branch of the NAFZ extending from the upper crust, where it has a width of ∼10 km, to the lower crust, where it widens to ∼30 km. This low‐velocity zone most likely extends into the upper mantle, where we constrain its width to be ≤ 50 km and interpret it as indicative of localized shear beneath the NNAF; this structure is similar to what has been observed for the NAFZ west of 32°, and therefore, we propose that the structure of the NNAF is similar to that of the NAFZ in the east. The southern branch of the NAFZ does not show a very strong signature in our images, and we conclude that it is most likely rooted in the crust, possibly accommodating deformation related to rotation of the Armutlu/Almacik Blocks situated between the two NAFZ branches.

Thermal Nature of Mantle Upwellings Below the Ibero‐Western Maghreb Region Inferred From Teleseismic Tomography

“…A number of teleseismic studies have measured variations in V P /V S by jointly inverting P and S arrival time data sets (e.g., Hammond & Humphreys, 2000;Schmandt & Humphreys, 2010a). A recent work by Papaleo et al (2018) estimated V P /V S from teleseismic data that constrain relative rather than absolute velocities; for this approach to be successful, accurate estimates of the background P wave and S wave velocity structure are required. However, both P wave and S wave velocity models are often not available for the same region and when they are, limitations arising from the resolving power of the tomography contaminate R S,P estimates (e.g., Deschamps & Trampert, 2003).…”

Revised supplementary material

“…Low Pwave velocities observed by Papaleo et al (2017) are also colocated with the low S-wave velocities detected in this study beneath the Pamukova basin. Papaleo et al (2017Papaleo et al ( , 2018 also found relatively high seismic velocity at depth within the Armutlu Block. We detect high S-wave velocities much closer to the surface that we attribute to the shallow metamorphic rocks reported in this region (Yılmaz et al, 1995).…”

“…Seismic velocity models of the crust in this region have also been constructed from teleseismic body wave tomography by Papaleo et al (2017Papaleo et al ( , 2018. They image depthaveraged seismic velocity between the surface and 90 km of depth, with a vertical and horizontal resolution of ∼ 15 km (Papaleo et al, 2017(Papaleo et al, , 2018.…”

“…Seismic velocity models of the crust in this region have also been constructed from teleseismic body wave tomography by Papaleo et al (2017Papaleo et al ( , 2018. They image depthaveraged seismic velocity between the surface and 90 km of depth, with a vertical and horizontal resolution of ∼ 15 km (Papaleo et al, 2017(Papaleo et al, , 2018. Despite the large difference in model resolution and a non-overlapping depth range, Papaleo et al (2017Papaleo et al ( , 2018 detect reduced crustal seismic velocities immediately to the north of the NAFZ, in the same regions we observe low S-wave velocities associated with the Adapazari basin, and heavily faulted sedimentary cover in the north-western part of the array (Figs.…”

“…If the regional anisotropy is primarily stress controlled, we would expect the seismic fast direction to be aligned in the direction of maximum compressive stress due to the preferential closure of fractures in this direction (Crampin and Lovell, 1991). However, Peng and Ben-Zion (2004) used local seismicity to show that the fast polarization direction at stations close to the ruptured Düzce fault ( Fig. 1) are generally parallel to and vary with the fault strike, suggesting an anisotropy mechanism determined by deformation fabric.…”