Title : The aseismic behavior of the North Anatolian fault : dynamics measured from multi-sensor InSAR time-series analysis and high-resolution optical images. Proposed co-advisors: Ziyadin Çakir (ITU), Cécile Lasserre, François Renard (ISTerre) Seismotectonic context, main objectives and method

par GÖkhan Aslan

Projet de thèse en Sciences de la Terre et de l'Univers et de l'Environnement

Sous la direction de Cécile Lasserre et de François Renard.

Thèses en préparation à Grenoble Alpes en cotutelle avec l'Istanbul Technical University , dans le cadre de Terre, Univers, Environnement , en partenariat avec Institut des Sciences de la Terre (laboratoire) depuis le 01-10-2015 .

  • Titre traduit

    Title : The aseismic behavior of the North Anatolian fault : dynamics measured from multi-sensor InSAR time-series analysis and high-resolution optical images.


  • Résumé

    The North Anatolian Fault (NAF) is one of the most active continental faults in the world. It accomodates the relative right-lateral motion between the Eurasia and Anatolia tectonic plates. During the past century, between 1939 and 1999, it ruptured in a unique westward propagating sequence of eight large destructive earthquakes, over a length of 800 km. The most recent Izmit and Dücze earthquakes in 1999 (Mw 7.6 and 7.2, respectively) broke the fault sections on the eastern side of the Marmara sea, a major transtensional stepover along the NAF, while the 1912 Ganos earthquake (Ms 7.4) ruptured its western side (partly within the basin). In between, a 70 km-long fault section below the Sea of Marmara remains unbroken, with a Mw 7.2 earthquake expected, putting a large threat on the city of Istanbul north of it. Two segments of the NAF inland exhibit aseismic creep within or below the seismogenic part of the crust that have been observed to interact with earthquakes. Creep at the base of the brittle crust near the hypocentral area of the Izmit earthquake was involved in the nucleation of the main shock. Conversely, shallow creep has been detected months to decades after large earthquakes, along the 1944 rupture and the supershear rupture of the 1999 Izmit earthquake. It may have initiated as postseismic, shallow afterslip decaying with time, before reaching a steady state. InSAR and GPS study of creep The variations of the degree of locking within the fault seismogenic zone, throughout the seismic cycle, depend on multiple parameters such as the geometry, lithological and mechanical properties of the fault zone, the past spatio-temporal distribution of earthquakes along the fault system, and the relaxation time versus the return time of earthquakes.  The main goal of this PhD project is to provide a fundamental understanding of the dynamics on these two creeping segments of the NAF and of how creep is related to seismicity and fault properties at various spatial and time scales. Determining the lateral variations and temporal characteristics (steady state or transient) of creep is of particular interest for the understanding of earthquake nucleation and arrest, and will be the main focus of the PhD work along the NAF. The goal is to combine GPS data and time series analysis of InSAR images acquired by various sensors at different wavelengths (TerraSAR-X, CosmoSkymed, Sentinel-1, ALOS1 and 2) as well as high-resolution optical images (a Digital Elevation Model -DEM- will be built from newly acquired stéréo Pléiades images). We will derive generic properties of creep and the distribution of fault frictional properties based on our past studies. The surface strain distribution in the near-fault zone will also be analyzed together with the precise morphology and segmentation of the fault (DEM analysis below), with the coseismic slip distribution of recent ruptures, and with potential seismic swarms or microseismicity activity recorded by the regional seismological network. Creep as seen from optical images ? The DEM built from Pléiades images, with metric ground resolution will be used for a detailed mapping and geological analysis of key fault zone outcrops sampled during the PhD of M. Kaduri. The geomorphological signature of creep will also be analyzed looking for specific morphologies of the fault segments at the surface (linear versus more complex segments, pull-aparts, and relay zones). Finally, we will also attempt to detect creep from correlation of optical images acquired since the 1950'. Image correlation methods allow for mixing of multi-platform images, either from satellite or from aerial photos. Hence, it allows looking at deformation over long time intervals. We will use archived aerial photo from the Turkish General Command of Mapping (1949 1955, 1972, 1974, 1983, 1984, and 2011 along Ismetpasa segment, 1999, 2006, 2009, 2012 along Izmit rupture) and early high-resolution satellite images correlated with new acquisitions by the satellite Pleiades (2014, 2015, pixel size of ~60 cm) to test the ability to measure creep along the North Anatolian Fault, and potentially its spatio-temporal evolution. The Insar and cGPS measurements performed in parallel will be used to benchmark the new developments required for the correlation methods. If the technics proves to be successful, the horizontal deformation field could be known with a ground spatial resolution of a meter or better, which would provide a new perspective to study the details of the geometry of the creeping section. Finally the results of creep data will be interpreted in terms of mechanical properties of the fault by comparing zones of creep with local geology (fault trace, rock composition) and inverting the surface displacements to get access to the dynamics of sliding at depth (locking depth, amount of coupling).