Dans cette thèse, nous présentons trois études concernant les interactions entre séismes et la phase de nucléation des forts tremblements de terre. La première partie de ce travail concerne le NO de la Turquie, où deux systèmes tectoniques sont présents : décrochement sur la Faille Nord Anatolienne et extension dans des essaims situés autour de la faille. Nous montrons que ces deux systèmes interagissent, mais répondent différemment à l'excitation du séisme d'Izmit (1999). Les calculs des changements de contraintes de Coulomb induits par le séisme d'Izmit indiquent que les évènements en décrochement répondent aux contraintes dynamiques, alors que les séismes en extension semblent contrôlés par la déformation statique de la croûte. Dans la seconde partie de cette thèse, nous analysons une importante séquence sismique qui s'est produite en 2008 en Grèce. Cette analyse a montré que les séismes en subduction ou en décrochement composant la séquence reflètent le mouvement du SO rigide de la Grèce vers le SSO accompagné du plongement de la plaque Africaine et du retrait vers le sud de la subduction. Cette première étape de déformation est suivie, quelques mois plus tard, de la déformation interne de la plaque Egée qui se manifeste par de l'extension N-S. Cette séquence sismique est le témoin de l'existence d'un couplage dynamique entre la subduction et le prolongement de la Faille Nord Anatolienne dans la région égéenne. Enfin, le but de la dernière partie de cette thèse a été d'apporter des éléments de réponse à la question : existe-t-il un phénomène observable par le biais de la sismicité qui précède de manière systématique les grands tremblements de terre? Pour cela nous avons étudié l'activité sismique précédant des séismes de M$geq 6.5$ dans des régions bien instrumentées (Japon, Cascades et NO des Etats-Unis). Nous montrons que la grande majorité des séismes interplaques est précédée d'une accélération de la sismicité, au contraire des séismes intraplaques. Nous mettons également en évidence une différence entre les séismes en décrochement et les séismes en subduction : les pré-chocs des séismes en subduction sont localisés dans une grande région, rendant un déclenchement en cascade des évènements et du choc principal impossible. / We present three studies about earthquake interactions and the nucleation phase of big earthquakes. The first part of this work deals with the northwestern Turkey, where two tectonic regimes exist, with the transform North Anatolian Fault and extension clusters around it. We show that these two systems interact, but respond differently to Izmit earthquake (1999) stimulation. Computations of the Coulomb stress changes induced by the Izmit earthquake indicate that strike-slip events respond to dynamic stresses, whereas the extension ones seem to be controlled by the static deformation of the crust. In the second part, we analyze a seismic sequence that occurred in 2008 in Greece. This analysis shows that subduction or strike-slip earthquakes of the sequence depict the overall motion of the southwestern Greece to the SSW, accompanied by the plunge and the southward retreat of the slab. This first deformation stage is followed, few months later, by the internal deformation of the Aegean plate, which occurs as N-S extension. This seismic sequence depicts the existence of dynamic coupling between the Hellenic subduction and the prolongation of the North Anatolian Fault. The aim of the last part was to shed some lights on the question: is there a systematic seismic phenomena preceding the occurrence of big earthquakes? To do that, we studied seismic activity preceding M$geq$6.5 earthquakes in some well-instrumented areas of the North Pacific. We show that most of the interplate earthquakes are preceded by an increase of the activity, unlike the intraplate ones. We also highlight a different behavior between strike-slip and subduction events: foreshocks of subduction earthquakes are located in a large area, preventing a cascade triggering.

Over the past century, a series of large (> 6.5) magnitude earthquakes have struck along the North Anatolian Fault Zone (NAFZ) in Turkey in a roughly East to West progression. The progression of this earthquake sequence began in 1939 with the Ms 8.0 earthquake near the town of Erzincan and continued westward, with two of the most recent ruptures occurring near the Sea of Marmara in 1999. The sequential nature of ruptures along this fault zone implies that there is a connection between the location of the previous rupture and that of the future rupture zones. This study focuses on understanding how previous rupture events and tectonic influences affect the stress regime of the NAFZ and how these stress changes affect the probability of future rupture along any unbroken segments of the fault zone using a two dimensional finite element modeling program. In this study, stress changes due to an earthquake are estimated using the slip history of the event, estimations of rock and fault properties along the fault zone (elastic parameters), and the far-field tectonic influence due to plate motions. Stress changes are not measured directly. The stress regime is then used to calculate the probability of rupture along another segment of the fault zone. This study found that when improper estimates of rock properties are utilized, the stress changes may be under- or over- estimated by as much as 350% or more. Because these calculated stress changes are used in probability calculations, the estimates of probability can be off by as much as 20%. A two dimensional model was built to reflect the interpreted geophysical and geological variations in elastic parameters and the 1939 through 1999 rupture sequence was modeled. The far-field tectonic influence due to plate motions contributed between 1 and 4 bars of stress to the unbroken segments of the fault zone while earthquake events transferred up to 50 bars of stress to the adjacent portions of the fault zone. The 1999 rupture events near Izmit and Düzce have increased the probability of rupture during the next ten years along faults in the Marmara Sea to 38% while decreasing the probability of rupture along the faults near the city of Bursa by ~6%. Large amounts of strain accumulation are interpreted along faults in the Marmara Sea, further compounding the case for a large rupture event occurring in that area in the future. / Ph. D.

Le but de cette thèse est centré sur la détection et la surveillance de la déformation de surface dans le nord-ouest de la Turquie, induite par une variété de phénomènes naturels (telles que l'activité tectonique, les glissements de terrain lents, etc.) et anthropiques (extraction des eaux souterraines, activités de construction, etc.), et sur l’analyse des mécanismes de déformation associés et de leurs conséquences pour l’environnement. Ce travail est basé sur le calcul de séries temporelles de déplacement du sol par interférométrie radar à synthèse d’ouverture (InSAR) afin d'analyser l'évolution des déplacements du sol, pour trois cas d’études associés à différents phénomènes géophysiques et processus sous-jacents. L’objectif de cette thèse est double : (1) révéler et quantifier les caractéristiques spatio-temporelles du glissement asismique le long de la rupture du séisme d’Izmit du 17 août 1999, et discuter de leur relation potentielle avec les propriétés de la faille (lithologie, géologie); (2) étudier la subsidence du sol dans des zones urbaines ou des zones exploitées par l'homme, induite par divers facteurs, et discuter des rôles relatifs de la tectonique, de la lithologie et des activités anthropiques dans ce mouvement du sol.Dans la première étude de cas, j’ai combiné des mesures InSAR, à partir d’images radar TerraSAR-X (bande X) et Sentinel-1 AB (bande C) acquises sur la période 2011-2017, à des mesures GPS en champ proche, effectuées tous les six mois à partir de 2014 jusqu'en 2016, ainsi qu’à des mesures de creepmeter, pour analyser le champ de vitesse en surface autour de la NAF, après le tremblement de terre d'Izmit de 1999. Les champs de vitesse moyenne horizontale InSAR révèlent que le taux de fluage (« creep ») sur le segment central de la rupture d'Izmit continue de décroître, plus de 19 ans après le séisme, ce qui concorde globalement avec les modèles de décroissance logarithmique des glissements post-sismiques de type « afterslip ». Le long de la section de la faille rompue à une vitesse « supershear » lors du séisme d'Izmit, le fluage se poursuit à une vitesse pouvant atteindre 8 mm / an. Un événement transitoire significatif, avec un fluage en accélération, est également détecté en décembre 2016 sur la série temporelle Sentinel-1, en accord avec les mesures d’un creepmeter installé près de la zone où la vitesse de fluage est maximum. Il est associé à un déplacement de surface total de 10 mm accumulé en un mois seulement.La deuxième cas d’étude porte sur l'identification et la mesure de la déformation du sol long-terme à Istanbul à partir d'une série InSAR couvrant près de 25 ans d'observations radar par satellite (1992-2017). Cette série temporelle InSAR a été calculée à partir d'images radar de plusieurs satellites (ERS-1, ERS-2, Envisat, Sentinel-1 A, B) afin d'étudier l'étendue spatiale et le taux de subsidence du sol dans la mégapole d'Istanbul.Dans le troisième cas d’étude, une série InSAR est calculée pour quantifier la subsidence de la plaine de Bursa (sud de la région de Marmara en Turquie), auparavant interprétée comme d’origine tectonique. Dans cette étude, StaMPS est utilisé pour traiter des séries d'images radar Sentinel 1 A-B acquises entre 2014 et 2017 le long d’orbites ascendantes et descendantse. Le champ de vitesse verticale obtenu après décomposition des champs de vitesse en ligne de visée sur deux traces complémentaires révèle que la plaine de Bursa s'affaisse à des vitesses allant jusqu'à 25 mm / an. A l’est, le signal de subsidence le plus important dans le bassin forme une ellipse allongée est-ouest et est limité par une plaine alluviale Quaternaire subsidant à environ 10 mm / an. Ces observations indiquent que l'accélération récente de la subsidence est principalement due aux activités anthropiques plutôt qu'aux mouvements tectoniques régionaux. / The aim of this thesis is centered on the detection and monitoring of surface deformation in northwest Turkey induced by a variety of natural (such as tectonic activity, slow moving-landslides, etc.) and anthropogenic (ground water extraction, construction activities, etc.) hazards and on the analysis of the related deformation mechanisms and their environmental consequences. In this work, I computed Interferometric Synthetic Aperture Radar (InSAR) time series to examine ground deformation evolution for three different case studies associated to different geophysical phenomena and underlying processes. The focus of this thesis is two-fold : (1) to reveal and monitor the spatio-temporal characteristics of aseismic slip along the August 17, 1999 Mw 7.4 Izmit earthquake rupture, and discuss its potential relationship with lithology and geology (2) to investigate ground subsidence in urban or human-exploited areas induced by various factors, and discuss the relative roles of tectonics, lithology and anthropogenic activities in such ground motion.In the first case-study, I combined InSAR measurements, based on X-band TerraSAR-X and C-band Sentinel-1 A-B radar images acquired over the period 2011-2017, with near field GPS measurements, performed every 6 months from 2014 to 2016, as well as creep meter measurements to examine the surface velocity field around the NAF after the 1999 Izmit earthquake. In this study, the Stanford Method for Persistent Scatterers InSAR package (StaMPS) was employed to process series of Sentinel 1 A-B (acquired along ascending and descending orbits) and TerraSAR-X (ascending orbits) radar images. The InSAR horizontal mean velocity fields reveal that the creep rate on the central segment of the 1999 Izmit fault rupture continues to decay, more than 19 years after the earthquake, in overall agreement with models of postseismic afterslip rate decaying logarithmically with time. Along the fault section that experienced a supershear velocity rupture during the Izmit earthquake, creep continues with a rate up to ~ 8 mm/yr. A significant transient event with accelerating creep is detected in December 2016 on the Sentinel-1 time series, consistent with creepmeter measurements, near the maximum creep rate location. It is associated with a total surface slip of 10 mm released in one month only. The second case study deals with the identification and measurement of secular ground deformation in Istanbul from a long-term InSAR time-series spanning almost 25 years of satellite radar observations (1992-2017). This InSAR time series was computed from radar images of multiple satellites (ERS-1, ERS-2, Envisat, Sentinel-1 A, B) in order to investigate the spatial extent and rate of ground subsidence in the megacity of Istanbul.In the third case study, InSAR time-series analysis is calculated for quantifying the subsidence of the Bursa Plain (southern Marmara region of Turkey), which has been interpreted as resulting from tectonic motions in the region. In this study, the StaMPS is employed to process series of Sentinel 1 A-B radar images acquired between 2014 and 2017 along both ascending and descending orbits. The vertical velocity field obtained after decomposition of line-of-sight velocity fields on the two tracks reveals that the Bursa plain is subsiding at rates up to 25 mm/yr. The most prominent subsidence signal in the basin forms an east-west elongated ellipse of deformation in the east, and is bounded by a Quaternary alluvial plain undergoing average vertical subsidence at ~10 mm/yr. The InSAR time series within the observation period is well correlated with changes in the depth of the ground water. These observations indicate that the recent acceleration of subsidence is mainly due to anthropogenic activities rather than tectonic motion.

Reintroduction of Anatolian mouflon population at Bozdag Protection &amp / Breeding Station to its former habitats(Emremsultan Wildlife Development Area in Ankara-Nallihan, and Karadag in Karaman) started in 2004. The magnitude of genetic change among Bozdag and reintroduced populations was evaluated by 11 microsatellite loci. Study populations revealed close results (&plusmn / st.dev.) &ndash / Bozdag population: nk = 2.9091 (&plusmn / 1.1362), AE = 2.0250 (&plusmn / 0.9537), Ho = 0.3830 (&plusmn / 0.2717), He = 0.3956 (&plusmn / 0.2746) / Nallihan population: nk = 2.9091 (&plusmn / 1.1362), AE = 2.0592 (&plusmn / 0.9451), Ho = 0.4086 (&plusmn / 0.2977), He = 0.4052 (&plusmn / 0.2767) / and Karadag population: nk = 2.5455 (&plusmn / 1.1282), AE = 1.8809 (&plusmn / 0.8758), Ho = 0.3388 (&plusmn / 0.2775), He = 0.3607 (&plusmn / 0.2716). Population differences for major genetic parameters were not significant (p &gt / 0.05) by comparisons with paired t-test. Also, temporal change in genetic diversity for Bozdag population was investigated by comparison with temporal data. Temporal changes in genetic parameters were found to be not significant and possible causes for differences were argued. Additionally, genetic diversity and PI computations for different traps were verified and compared to uncover any potential bias due to the catching method. Comparisons did not reveal significant differences illustrating the homogeneity among traps. On the other hand, simulations detected the higher sensitivity of allelic diversity (A) to founder events than P and heterozygosity (Ho &amp / He) levels which supports heterozygosity excess method for bottleneck analysis. With the same simulation analysis, observed genetic diversity within reintroduced samples were found to be in the ranges of expectation (99% CI) indicating that translocated individuals were chosen randomly. Bottleneck analysis based on heterozygosity excess method (one-tailed test for heterozygosity excess: pSMM = 0.28515, pTPM = 0.06445, pIAM = 0.02441) and allele frequency distributions method (normal L-shaped) could not detect a recent genetic bottleneck for Bozdag population. However, simulations determined that these two methods are prone to type II error. Bottleneck detection failure for the study population is probably due to type II error instead of other sources of error like violations of model assumptions.

The Anatolian mouflon (Ovis gmelinii anatolica) is an endemic ungulate subspecies and of IUCN Vulnerable status that inhabits Konya-Bozdag region located in Central Anatolia. In this thesis, the demography and habitat use of the only natural population of Anatolian mouflon at Konya-Bozdag Provice is studied. Throughout the study, distance sampling techniques, specifically line transect sampling, are used to estimate density, size and growth rate of this population. Sex, age and count data are used to estimate relevant parameters. Changes in population structure and seasonal area use are monitored for two years. Data is collected during repeated random line transects from May 2007 to July 2009. Line transect method is preferred since it requires less effort and is less expensive when compared to the complicated techniques that need animal marking or radio tagging. In total, 78 transect lines were surveyed during the time which covers 3 lambing periods. In spite of the paratuberculosis epidemics in the fenced area which has affected the population adversely in the previous years, a rather stable population trend is observed. The post-breeding population size in 2007, 2008 and 2009 were estimated to be 883+-241, 939+-136, 972+-243 (average+-satndard error) and densities as 27.227, and 28.186 individuals per sq km, respectively. Growth rate of population is found using the ratio of the population size estimates of consecutive years from 2007 to 2009, average 1.0495+-0.0203. Habitat use patterns of the Anatolian mouflon throughout the study period are investigated according to seasons and sex groups. There is sexual segregation in the Anatolian mouflon population in Konya Bozdag Province, with the females using the western part and males using the eastern part of Bagderesi. Seasonal patterns affect on the area use of animals, group formation and compositions and the relationship between them are searched throughout the study. The movements of the individuals and groups followed seasonal patterns as centers of activities changed according to seasons.

&lsquo / Bag&rsquo / settlements have been a part of dynamic spatial systems, which seasonally depend on and reciprocally interact with Anatolian towns throughout centuries. This thesis presents the transforming setting of &lsquo / bag&rsquo / settlements related to changing values and meanings through an ontological assessment. Therefore, the thesis assumes that the main values of spatial organization, farmland pattern, ecological formation, settlement character, socio-cultural structure and lifestyle of the inhabitants constitute the unique entity of &lsquo / bag&rsquo / settlements. In time, the pressures of changing socio-economic conditions have destroyed the interaction between Anatolian towns and &lsquo / bag&rsquo / settlements. The thesis explains the changing role of &lsquo / bag&rsquo / settlements, changing and conflicting land uses and the loss of unique &lsquo / bag&rsquo / character in the case of Mugla-Karabaglar. Karabaglar is a &lsquo / bag&rsquo / settlement in the southwestern Turkey, where Mugla town residents live seasonally. It is a third grade natural site, the unique character, the natural and cultural assets and the lifestyle of which must be preserved. However, with transformation of the main values, meanings and practices, Karabaglar could no longer perpetuated its initial existence of being. Karabaglar is significant for two reasons: it has natural and cultural beings that need to be conserved, and as it still goes through a transformation process. The master&rsquo / s thesis of the author evaluated conservation plan of Karabaglar. Differently, this doctoral thesis evaluates the dynamics of the spatio-temporal transformation process in Karabaglar. It presents the changing role and significance of Karabaglar within the town-country continuum. This situation brings forward the conservation problematic of the character, landscape and uniqueness of Karabaglar. Within this respect, the thesis contributes to the literature of &lsquo / bag&rsquo / settlements in terms of defining their being and changing role throughout the history.

The Hatay Triple Junction (HTJ) cannot be described by a simple model with three major plates as proposed by previous studies. A more complex block model is proposed in this study by adding the Iskenderun block and Amanous micro block, the Karasu fault and Karatas-Osmaniye fault being defined as individual faults not as the extension of other major faults in the region. Our modeling assumes that the Maras triple junction is formed by the connection of the Karatas-Osmaniye Fault (KOF) with the Karasu Fault (KF) and the East Anatolian Fault (EAF). The KF shows a sinistral slip rate of 4.0±1.0 mm/yr and a compressional behavior with a compression rate of 2.1-2.7 mm/yr which contradicts the extensional nature proposed by previous studies. The EAF shows pure left lateral slip rate of 9.0±0.3 mm/yr with no significant extension or compression; the DSF has a slip rate of 3.5±0.3 mm/yr over the northern and southern segments; the KOF has a 3.6±0.7 mm/yr; the Cyprus arc has a clear compressional deformation with a revers slip rate of 2.0-5.0 mm/yr and with no significant strike-slip component. The relative Euler poles are estimated in this block modeling, we define the Anatolia-Arabia Euler pole at (27.61±0.98 °N, 45.127±2.45 °E, 0.391± 0.056°/Myr), and (31.012±1.51 °N, 46.464±4.44 °E, 0.202±0.067°/Myr) Sinai-Arabia Euler pole.

The purpose of this thesis is to test the validity of applications of Remote Sensing and Geographical Information Systems in Anatolian archaeology. The focus of the study is an Iron Age fortress &Ccedil / evre Kale and its associated structures. During the course of the study, 5 km long outer wall enclosing a territory around &Ccedil / evre Kale documented for the first time by employing high altitude aerial imagery. In addition to the GIS analyses, examination of the geology, land use and soil quality data showed that the outer wall is in a way acting to guard and protect inhabitants of the fortress and, perhaps more importantly, the well-watered pasture surrounding the fortress and demarcated by the enclosure wall. Evaluation of the available archaeological and historical evidence suggested that &Ccedil / evre Kale might be of a site with significant military importance at least in the first half of the 6th century BC. As a result, this thesis is underlying the importance of high and low altitude aerial imagery in terms of documentation, evaluation and monitoring of the archaeological sites as part of the archaeological research

Gemlik County is located in the Marmara Region (NW Turkey), which has been affected by destructive earthquakes sourced from North Anatolian Fault System throughout its history. The bulk of the settlement rests on alluvial deposits of the Gemlik pull-apart basin. So, it is vital to investigate the foundation soils in this basin and the response of them to earthquakes. Many earthquake codes were established by the authorities in different countries of the world to estimate the possible ground shaking and seismic loads which may act on buildings. In this study, Turkish Earthquake Code (TEC-1998) and Eurocode-8 (EN-1998) have been utilized. The analyses showed that EN-1998 results in more conservative estimates relative to TEC-1998, in terms of spectral ordinates. The source of difference between TEC-1998 and EN-1998 has been investigated and three possible reasons have been identified. The variation is probably due to the different seismic characteristics of Turkey and Europe, different soil amplification levels defined by the codes, and different soil classification procedures of the codes.

L’histoire de l’Anatolie « hittite » constitue une page importante de l’histoire du Proche-Orient ancien du deuxième millénaire avant J.-C. Pendant longtemps après la découverte des Hittites la recherche s’était focalisée sur le cœur de l’Empire hittite. Excepté le Sud-Est les zones périphériques sortent peu à peu de l’ombre C’est dans cette perspective que cette thèse s’inscrit : les rapports entre l’empire hittite et les entités politiques situées à l’Est dont le territoire avait été considéré comme le berceau d’un peuple d’origine toujours obscure : les Arméniens.Un certain nombre d’inscriptions cunéiformes hittites ont été étudiés dans le but de trouver des traces des Proto-Arméniens en Anatolie orientale au deuxième millénaire avant J.-C. Notre recherche est donc constituée selon le plan suivant : Dans le cadre du premier chapitre, nous examinerons l’histoire de la philologique régionale, y compris les langues. Nous présenterons les écoles de pensée et les méthodes appliquées par les grands savants occidentaux et soviétiques dès 19ème siècle pour le classement de ces langues « voisines » régionales qui appartiennent pourtant à des familles linguistiques différentes.Dans le cadre du deuxième chapitre, nous étudierons la géographie historique de l’Anatolie orientale (régions mentionnées par les sources hittites et arméniennes), considérées comme le berceau des Arméniens. Le troisième chapitre a été consacré à l’étude comparée du vocabulaire hittito-arménien. Dans le cadre du dernier chapitre (chapitre 4), nous présenterons et étudierons les textes hittites relatifs à l’Anatolie orientale. / The history of 'Hittite' Anatolia is an important part of the history of the ancient Near East of the second millennium B.C.For a long time after the discovery of the Hittites, the research had focused on the heart of the Hittite Empire. With the exception of the Southeast, the peripheral areas gradually emerge from the shadows. It is in this perspective that this thesis takes place: the relations between the Hittite empire and the political entities situated in the East, whose territory had been considered the cradle of a people of obscure origin : the Armenians.A number of Hittite cuneiform inscriptions have been studies with tha aim of finding traces of the Proto-Armenians in Eastern Anatolia in the second millennium B.C.Our research constituted according to the following plan: Within the framework of the first chapter, we will examine the history of regional philology, including languages. We will present the schools of thought and the methods applied by the great Western and Soviet scholars of the 19th century for the classification of these neighboring regional languages which belong to different linguistic families.In the second chapter, we will study the historical geography of Eastern Anatolia (areas mentioned by Hittite and Armenian sources), considered the cradle of the Armenians.The third chapter was devoted to the comparative study of the Hittite-Armenian vocabulary. In the last chapter (chapter 4), we will present and study the Hittite texts relating to Eastern Anatolia.