Knowledge of earthquake source faults is crucial to calculate robust Coulomb stress models. However, source faults are often poorly constrained, especially for pre-instrumental events, and these historic earthquakes are commonly studied with little or no consideration of other nearby events. We introduce an approach using Coulomb Stress Transfer (CST) modelling to investigate historical earthquake sequences and constrain possible source faults. Using historical and instrumental records from western Turkey, we create an ensemble of earthquake sequences featuring multiple rupture scenarios for individual earthquakes, and model both coseismic and interseismic CST. We filter and evaluate the models based on criteria to gain knowledge on historic earthquakes and their source faults and assess the current stress state and related seismic hazard of the investigated fault network. The results provide further constraints on the source faults of several historic earthquakes, and therefore can be used to refine historic earthquake studies from a novel perspective.
Quantifying interseismic deformation of fault networks which are predominantly deforming in a north-south direction is challenging, because GNSS networks are usually not dense enough to resolve deformation at the level of individual faults. The alternative, synthetic aperture radar interferometry (InSAR), provides high spatial resolution but is limited by a low sensitivity to N-S motion. We study the active normal fault network of Western Anatolia, which is undergoing rapid N-S extension, using InSAR. In the first part of this study, we develop a workflow to assess the potential of decomposing InSAR line-of-sight (LOS) velocities to determine the N-S component. We use synthetic tests to quantify the impact of noise and other velocity components and outline the requirements to detect N-S deformation in future studies. In its current state, the N-S deformation field is too noisy to allow robust interpretations, hence in the second part we complement the study by including vertical deformation. Since most faults in the study region are normal faults, the high-resolution vertical velocity field provides new insights into regional active faulting. We show that tectonic deformation in the large graben systems is not restricted to the main faults, and seemingly less active or inactive faults could be accommodating strain. We also observe a potential correlation between recent seismicity and active surface deformation. Furthermore, we find that active fault splays causing significant surface deformation can form several kilometres away from the mapped fault trace, and provide an estimate of current activity for many faults in the region.