Elastic and inelastic triggering of earthquakes in the North Anatolian Fault zone

Abstract

Deformation models used for explaining the triggering mechanism often assume pure elastic behaviour for the crust and upper mantle. In reality however, the mantle and possibly the lower crust behave viscoelastically, particularly at long-term scale. Consequently, the stress field of an earthquake is in general time dependent. In addition, if the elastic stress increase were enough to trigger a later earthquake, this triggered event should occur instantaneously and not many years later. Hence, it is adequate to include inelastic behaviour when analysing stress transfer and earthquake interaction.In this work, we analyse a sequence of 10 magnitude M s > 6.5 events along the North Anatolian Fault between 1939 and 1999 to study the evolution of the Coulomb stress field.We investigate the triggering of events in the series by stress transfer, taking viscoelastic relaxation into account. We evaluate the contribution of elastic stress changes, of post-seismic viscoelastic relaxation in the lower crust and mantle, and of steady tectonic loading to the total Coulomb stress field. We analyse the evolution of stress in the region under study, as well as on the rupture surfaces of the considered events and their epicentres. We study the state of the Coulomb stress field before the 1999 Izmit and Düzce earthquakes, as well as in the Marmara Sea region.In general, the Coulomb stress failure criterion offers a plausible explanation for the timing and location of the events in this area and for this sequence of events. However, we show that using a purely elastic model disregards an important part of the actual stress increase/decrease. In several cases, post-seismic relaxation effects are important and greater in magnitude than the stress changes due to steady tectonic loading. Consequently, viscoelastic relaxation should be considered in any study dealing with Coulomb stress changes. of 39According to our study, and assuming that an important part of the rupture surface must be stressed for an earthquake to occur, the most likely value for the viscosity of the lower crust or mantle in this region is 5⋅10 17 -10 18 Pa⋅s. Our results also suggest that other time-dependent processes apart from viscoelastic relaxation might have been involved in the triggering of the 1999 Düzce event. Finally, we argue that the Marmara Sea region is currently being loaded with positive Coulomb stresses at a much faster rate than would arise exclusively due to steady tectonic loading on the North Anatolian Fault.

Seismic behaviour of the North Anatolian Fault beneath the Sea of Marmara (NW Turkey): implications for earthquake recurrence times and future seismic hazard

“…These studies have comprised reassessment of historical seismicity (Ambraseys and Jackson 2000;Ambraseys 2001aAmbraseys , b, 2002a, monitoring of microseismicity (Sato et al 2004), high-resolution bathymetry and multi-channel seismic reflection studies (Okay et al 2000;Le Pichon et al 2001;Armijo et al 2002Armijo et al , 2005Parke et al 2002;Gökaşan et al 2003;Demirbag et al 2003;McHugh et al 2006;Kanbur et al 2007), earthquake focal mechanism-derived stress orientations (Polat et al 2002;Pınar et al 2003), field geology (Altınok et al 2003;Seeber et al 2004;Okay et al 2004;Altunel et al 2004), paleoseismology (Rockwell et al 2001;Klinger et al 2003;McHugh et al 2006), Coulomb static stress change modelling (Hubert-Ferrari et al 2000;Parsons et al 2000;Utkucu et al 2003;Parsons 2004;Lorenzo-Martin et al 2006;Pondard et al 2007), space geodesy (Meade et al 2002;Flerit et al 2003) and earthquake probability (Parsons et al 2000;Parsons 2004;Cisternas et al 2004) studies. The basin of the Sea of Marmara was considered as a rift with a strike parallel to the proposed western Turkey rift system in the earlier phase of the construction of the plate kinematics of the eastern Mediterranean region (McKenzie 1972).…”

Coulomb stress transfer and accumulation on the Sagaing Fault, Myanmar, over the past 110 years and its implications for seismic hazard

“…The tectonic stress loading was modeled by the deep dislocation technique proposed by Savage [] and was realized by a steady slip from the locking depth to 100 km. The slip increases from zero at the locking depth to its full magnitude at the bottom of crust [ Lorenzo‐Martín et al , ]. In our simulation, the locking depth was set to be 15.5 km (bottom of upper crust), which is corresponding with the result from a GPS study in the central Myanmar [ Vigny et al , ].…”

Elastic and visco-elastic stress triggering in the South Iceland Seismic Zone due to large earthquakes since 1706

“…no E-W trending through-going fault, but an en-echelon set of N-S trending faults), the application of stress transfer analysis techniques (horizontal shear stress and Coulomb stress changes) seems to be inappropriate for time intervals of more than a few days. However, the modelling approach presented here might be used elsewhere (provided input data quality is good) to identify future regions of potential earthquake hazard and to mitigate earthquake risk, as shown by Lorenzo-Martín et al (2006). for shear stress evolution, centring the background field at 64.0° N. The stress field was evaluated before and after each event.…”