Measuring Anatolian plate velocity and strain with Sentinel-1 InSAR and GNSS data: implications for fault locking, seismic hazard, and crustal dynamics

Abstract

<p>Geodetic measurements of small rates of interseismic crustal motion made at high spatial resolutions and over large areas are crucial for understanding the earthquake cycle, characterizing spatial variations in lithosphere rheology and fault frictional properties, illuminating the mechanics of large-scale continental deformation, and improving earthquake hazard models. The densification of Global Navigation Satellite System (GNSS) networks has provided an unprecedented view of the kinematics of deforming regions. However, large gaps in spatial coverage still hamper our ability to fully characterize patterns of surface deformation. Interferometric synthetic aperture radar (InSAR), and the Sentinel-1 (S-1) satellites in particular, have the potential to overcome this obstacle by providing spatially continuous measurements of surface motions, without instruments on the ground, with precision approaching that obtained from GNSS, and at a resolution of a few tens of meters. In order to manage and process the large data volumes produced by S-1, we have developed open-source, automated workflows to efficiently generate interferograms and line-of-sight (LOS) velocity fields. These outputs are valuable for a range of applications, from earthquake rapid-response to investigating human-induced ground-level changes. In this talk, we demonstrate our ability to measure plate-scale interseismic deformation using data from the first ~5 years of the S-1 mission. We estimate LOS velocities for the Anatolian microplate, an area encompassing ~800,000 km<sup>2</sup> and including the highly seismogenic North Anatolian Fault Zone (NAFZ). By combining S-1 InSAR and GNSS data, we create high-resolution surface velocity and strain rate fields for the region, which illuminate horizontal deformation patterns dominated by the westward motion of Anatolia relative to Eurasia, localized strain accumulation along the NAFZ, and rapid vertical signals associated with groundwater extraction. Relatively low levels of strain characterize other active regions including the East and Central Anatolian Fault Zones and the Western Anatolian Extensional Province. We find that GNSS data alone are insufficient for characterizing key details of the strain rate field that are critical for understanding the relationship between strain accumulation and release in earthquakes. We highlight two important results stemming from our work including probabilistic estimates of the recurrence times of earthquakes of varying magnitudes for the region and a new NAFZ locking distribution that shows close correspondence to the surface rupture extents of large 20<sup>th</sup> century earthquakes.</p>