Determination of aseismic tectonic structure in Bolu-Çankiri and Amasya regions through the North Anatolian fault(NAF) with periodic GPS observations

Earth crust is in a continuous motion due to the convection current in the mantle layer below it and divided into several parts in time. These structures' movements have played the most important role in the formation of continents and called "tectonic plates". These movements can only be observed with modern observation techniques and still in motion for millions of years. Hence, fault zones are formed at the block boundaries. These tectonic blocks move relative to each other, and this generally reveals with 4 different movement types: If the density of two neighbor blocks are close or equal to each other, then they overlap on each other to form depression. However, if the density of blocks are different, then the lower density block *dives under the other. Also, seperated blocks move apart from each other because of the tectonic mechanism. And the last type is observed between the blocks which have lateral movements between them. Fault zones are the regions which relative block movements attenuated. In these regions, blocks don't have free motion due to the friction force in the rocks, they're partially or fully locked. While these blocks continue their movements, the rocks along the fault resist this movement to a certain threshold which causes strain accumulation. When this strain prevail the friction force, current potential energy is emerged. This energy discharge is called "earthquake". Most of the faults on the earth crust have the potential to produce earthquakes periodically. Block movements and determination of the accumulated strain along the faults are important factors for the assessment of potential earthquake risk. In addition to seismic and geological data, different techniques are also available in the field observations. These motions can be determined with high precision with the help of GPS, InSAR, LIDAR and creepmeter measurements. In addition, periodic or continuous observations on the stations, which established around or nearby the fault zone, can provide data for seismic risk assessment. Experimental studies shows that observation at the stations near and far from the fault plane can be used to predict the locking depth of the fault, and distant stations are suitable for long-term block movement detection. GPS is the most prominent method along these studies. It has the advantage to provide low cost, portable receivers, high precision and flexible network design on the field and has been used since the 90s. The geography of Anatolia hosts continuous block movements and significant faults in the terms of seismic risk. Most of our country is located on the Anatolian block, which is compressed by a much bigger Eurasian block on the north and northward movement of the Arabian block on the south, and moves to the west ~20 mm/year relative to the Eurasian block. These significant movements results two important fault zones over time: North Anatolian Fault Zone(NAFZ) and East Anatolian Fault Zone(EAFZ). NAF is a right lateral strike slip fault, which starts from Karlıova in the East and passes through mid Anatolia by Erzincan, Ladik, Kargı, Ismetpasa, Gerede, Izmit and reaches the Marmara and north of Eaegan Sea with two branches. This almost 1200 km structure causes many destructive earthquakes in the history. Studies indicates that these earthquakes migrate from the east to the west along the fault. This phenomenon starts with 1939 Erzincan earthquake, and finally breaks the NAF into the Marmara Sea with 1999 Gölcük and Düzce earthquakes. Most of the fault on the earth surface are locked and zero movement observed in these block boundaries, fault planes, due to friction between rocks. The structure of the NAF demonstrates this general situation and is generally locked to the bottom of seismogenic layer(15-20 km depths). However, a few specific region on earth slip freely without an earthquake and can be observed from the fault surface. These rare events are defined as "aseismic creep", where NAF also hosts this phenomena on the middle part. Aseismic creep can be observed to the bottom of seismogenic layer or to some certain depths. In the first case, an earthquake may not occur if the creep ratio is close or equal to the long-term block movements in the region. In the second case, if the aseismic creep is not occurred to the bottom of seismogenic layer, or if this movement's speed is lower than the long-term block's relative motions, the accumulation of the energy for an earthquake cannot be prevented. But in any case, this rare nature event should be monitored carefully for seismic risk assessment. There are two segments on the NAF reported to have this rare movement on its route to the Eagean Sea: Ismetpasa and Destek. In these regions, NAF is not fully locked and have some deformations which can be observed from the earth surface and directly related to the geological structure of the fault plane in that region. Aseismic creep in Ismetpasa segment, which is almost 100 km to the north of Ankara, is monitored since this movement has been observed on a wall at 1970. Until then numerous studies were conducted in the region to better understand the mechanism and creep ratio with different techniques such as Geodetic (GPS, InSAR and terrestrial measurements) and Geological and Geophysical methods. Geodetic surveys and studies generally focused on the Ismetpasa town and on Hamamlı village, a few kilometers to the east of Ismetpasa, where GPS, creepmeter, LIDAR, etc. techniques are used. These studies cover only a few kilometers of the whole segment. Thus, other InSAR based studies also carried out in the region and revealed that aseismic creep occur almost 70-80 km around the segment. Results from these studies show different creep velocities, and they are not suitable to detect the 3D rate of this rare event. On the other hand, aseismic creep has been observed recently at the Destek segment, which is almost 30 km to the northwest of Amasya, and there is not much studies around this segment. NAF have many earthquakes on the last century which effects the aseismic creep along the segments. There seems that 3 main earthquakes occurred along these regions before the catastrophic 1999 earthquakes: Ladik-Kargı-Tosya-Ilgaz 1943 earthquake (Ms=7.2), Bolu-Gerede 1944 earthquake (Ms=7.2), and Kursunlu 1951 earthquake (6.9). Studies around this region cannot clearly reveal that if this creeping movement starts before, after or between these earthquakes. In this context, it was evaluated that there were not enough observation stations along the NAF's aseismic creep regions, insufficient GPS data or velocity area. For this purpose, a GPS network established on the field and all stations of it have been observed between 2014-2016 by a large scaled network. There were 5 profiles planned along the study area: 4 on the Ismetpasa, and 1 on the Destek segment. Through the field studies, 23 points established, where 9 of them are pillars and the rest of them are placed on the local rocks. There is another two-point profile between 2nd and 3rd profiled formed because of the suitable positions of exiting pillars. In addition, a new point established on the 2nd profile at the intersection between the profile and fault plane. Finally, there are 23 campaign stations were established, but one of the stations damaged after the observations on 2014. It is removed from the further evaluation and a 22 station network has been observed. Also, data of 30 continuous GPS stations around the region (CORS-TR) are evaluated between 2009-2016, to make connection between global network and Ismetpasa -Destek and to strengthen to velocity field between them for block modelling. This allows to form a network including 52 stations. Velocities for these two segments and the segment between them were estimated with the GAMIT/GLOBK software at first, and all velocities for the observation stations are calculated at the ITRF2008 reference frame where Anatolian plate moves westwards relative to a fixed Eurasian plate. The result of this step were used as input to the block modelling with the TDEFNODE software. Earlier studies along these segments and seismic activities examined carefully to reveal the model parameters for each segment. Finally, the fault plane geometry estimated in these regions and seismic risk has been evaluated according to the creep movement, and all the steps of the processes explained in detail. Model results indicate that the creeping event continues at both segments since they discovered. Fault plane in both segments show aseismic creeping event to some depths, and attenuate almost half way to the bottom of seismogenic layer. Creeping trend at both segment seems to accelerate, but much clear at the Destek segment. Nevertheless, this movement cannot prevent a large-scale earthquake in the mid-long term, thus the fault plane is not fully locked and the creeping ratio at both segments are lower than the long-term relative block movements.