Characterization of the seismic velocity field and crustal anisotropy in the Marmara region, NW Turkey, using ambient noise
Geophysics
Final Report Abstract
The initial aim of this project was to study the ambient noise field at three selected locations of the North Anatolian Fault Zone (NAFZ) in NW Turkey to determine the near-surface and upper crustal S-wave velocity structure as well as seismic anisotropy and its spatiotemporal variations during the seismic cycle from continuous passive seismic recordings. We intended to focus on three selected spots of the NAFZ representing different stages during the seismic cycle. For reasons such as data availability, the increasingly relevant earthquake hot-spot offshore Istanbul and in particular the occurrence of the M7.1 Van earthquake in eastern Anatolia in 2011 we decided to focus on two key locations along the NAFZ: These are the eastern Marmara region offshore Istanbul where a M>7 earthquake is pending, and the Van region in eastern Turkey to study the pre-/co- and postseismic velocity field framing the 2011 M7.1 earthquake. The ultimate goal of this project was to determine the uppermost crustal velocity field and its spatial and temporal variations which was successfully achieved for both regions as outlined in more detail in the report below. The main results were that we directly measured a co-seismic change (drop) of seismic velocity in the vicinity of the M7.1 2011 Van/eastern Turkey earthquake using ambient noise recordings. The broadband waveform recordings were cross-correlated and stacked for a time period of six months framing the mainshock allow determining an average Rayleigh-wave velocity of 3.3 km/s for the target region. The observed co-seismic velocity decrease in the hypocentral region reached up to ~0.8 % in the frequency band of 0.05 – 0.3 Hz. The observed velocity drop is largest close to the earthquake hypocenter and decreases with distance from the mainshock rupture. A similar correlation was observed between the co-seismic velocity decrease and the amount of co-seismic slip on the part of the rupture plane penetrated by the corresponding ray paths. Within the frequency band of 0.05-0.08 Hz, corresponding to the Van hypocentral depths, an increase in velocity of about ~0.50 % was observed outside the high-slip area, indicating an increase in stress level. The observation of a drop in crustal velocity in conjunction with a M7+ earthquake shows a clear response of the brittle crust pointing out a change in elastic parameters. Such changes can be detected and even monitored with conventional regional seismic networks with inter-station spacing of several tens of kilometers and by applying state-of-the-art ambient noise processing techniques. In the second study we successfully applied a cross-correlation analysis of ambient noise surrounding the eastern Sea of Marmara region. We find that at least 60 days of waveform recordings are required to obtain reliable Green’s functions, with this station geometry. We observed that the Thrace Basin in the northern Istanbul Peninsula hosts a rather lower velocity structure compared to the rest of the crustal block and also that on-shore crustal blocks surrounding the eastern Sea of Marmara have similar (higher) seismic velocities while the Çınarcık Basin in between reflects lower surface wave velocities as can be expected. This study has defined the sedimentary and upper crustal S-wave velocity structure of the eastern Sea of Marmara region. Furthermore, using an average dispersion curve obtained in this study, we determined a 1-D S-wave velocity model for the target area, which increases the hypocenter determination precision for seismicity along the Princes’ Islands segment of the NAFZ offshore of İstanbul from ambient noise recordings. The third study focused on determining anisotropy in the eastern Sea of Marmara region based on additional broadband seismic stations of the area. Interestingly, a large difference between Love and Rayleigh dispersion curves was identified if correlation paths crossed the fault zone. So far, the ray coverage is not sufficient to identify azimuthal anisotropy. However, Love-Rayleigh-discrepancy can still be observed. This discrepancy between Love and Rayleigh wave group velocity dispersion curves may indicate a structural anisotropy. The reliability of these results needs to be investigated further.
Publications
- (2014). Ambient Noise Analysis in the Eastern Sea of Marmara Region in Northwest Turkey: Lateral Variations of the Crustal Velocity Field. Bull. Seismol. Soc. Am., 104(4), 1954-1963
Acarel, D., Bulut, F., Bohnhoff, M.
(See online at https://doi.org/10.1785/0120130160) - (2014). Co-seismic velocity change associated with the 2011 Van Earthquake (M7.1): Crustal response to a major event. Geophys. Res. Lett., 41, 4519-4526
Acarel, D., Bulut, F., Bohnhoff, M., Kartal, R.
(See online at https://doi.org/10.1002/2014GL060624) - (2015). Characterization of the crustal velocity field in space and time using seismic ambient noise. PhD thesis, Freie Universität Berlin, 2015
Acarel, D.
- (2017). GONAF–A borehole Geophysical Observatory around the North Anatolian Fault in the Eastern Sea of Marmara, Sci. Dril., 5, 1–10
Bohnhoff, M., Dresen, G., Ceken, U., Kadirioglu, F.T., Kartal, R.F., Kilic, T., Nurlu, M., Yanik, K., Acarel, D., Bulut, F., Ito, H., Johnson, W., Malin, P.E., Mencin, D.
(See online at https://doi.org/10.5194/sd-22-19-2017)
