Earthquake cycle of the North Anatolian fault along the rupture zone of the august 17, 1668 great Anatolian earthquake

In this study, we investigated the earthquake cycle along the rupture zone of the August 17, 1668 Great Anatolian Earthquake (M8.1) combining GPS measurements and earthquake records. The 1668 M 8.1 earthquake ruptured 450 km section of the NAFZ. Based on the historical and paleo earthquake records, we detected three complete and one incomplete earthquake cycles since 1254. We elaborated on elastic rebound theory investigating the creeping and the locked stages of the individual fault segments along the NAFZ. We combined all available GPS measurements and modeled them using arctangent approach to simultaneously estimate segment-based slip rates and locking depths. Slips rates are then used to estimate preliminary inter-seismic slip storages assuming fully locked fault segments right after the mainshocks. The misfits between co-seismic slips and preliminary inter-seismic slip storages of historical earthquake indicate that the fault does not store slip for a while after a major earthquake. Our analysis showed that there is a partitioning of the creeping and locked stages both in time and space. Only along one the segments, the 1943 (M7.7) rupture zone, exceptionally creep played a minor role during the seismic cycle (0.1%). Along the 1939 (M7.9), 1957 (M7.0), 1967 (M7.2), and 1999 (M7.5) rupture zones, creep played a considerable role (16.9%, 22.2%, 17.9% and 22.4%, respectively). Along the 1942 (M7.1), 1944 (M7.4), 1999 (M7.1) rupture zones, creep played a substantial role, and covered almost half of the seismic cycle (54.4%, 44.0% and 48.3%, respectively). The segments host currently different earthquake potentials as they have distinctive creeping/locking rates despite the fact that they are exposed to similar deformation rates (ranging between 19.5±0.5 – 24.2±0.3 mm/y). In this frame, the NAFZ is currently able to generate M7+ earthquakes along the 1939 (M7.9), 1943 (M7.7), and 1944 (M7.4) rupture zones, M6.5+ earthquakes along the 1942 (M7.1), 1957 (M7.0), 1967 (M7.2), and 1999 (M7.5) rupture zones, and an M6+ earthquake along the 1999 (M7.1) rupture zone. Our results show that the GPS-derived slip rates systematically accelerate from the east to the west. GPS-derived locking depths do not demonstrate a systematic trend. They are deeper along the central NAFZ and shallower along the other sub-segments. Failure of the NAFZ will probably end at the western segments within 239±3 years. The space-time pattern of the earthquakes during the last three complete and the current incomplete cycles confirms that the failure of the NAFZ starts from the east, and the earthquakes systematically migrate to the west outlining a decelerating behavior.