Weak, Seismogenic Faults Inherited From Mesozoic Rifts Control Mountain Building in the Andean Foreland

New earthquake focal mechanism and centroid depth estimates show that the deformation style in the forelands of the Andes is spatially correlated with rift systems that stretched the South American lithosphere in the Mesozoic. Where the rifts trend sub-parallel to the Andean range front, normal faults inherited from the rifts are being reactivated as reverse faults, causing the 30--45 km thick seismogenic layer to break up. Where the rift systems are absent from beneath the range front, the seismogenic layer is bending and being thrust beneath the Andes like a rigid plate. Force-balance calculations show that the faults inerhited from former rift zones have an effective coefficient of static friction < 0.2. In order for these frictionally-weak faults to remain seismogenic in the lower crust, their wall rocks are likely to be formed of dry granulite. Xenolith data support this view, and suggest that parts of the lower crust are now mostly metastable, having experienced temperatures at least 75--250 degrees hotter than present. The conditions in the lower crust make it unlikely that highly-pressurised free water, or networks of intrinsically-weak phyllosilicate minerals, are the cause of their low effective friction, as, at such high temperatures, both mechanisms would cause the faults to deform through viscous creep and not frictional slip. Therefore pre-existing faults in the Andean forelands have remained weak and seismogenic after reactivation, and have influenced the style of mountain building in South America. However, the controls on their mechanical properties in the lower crust remain unclear.

New seismological data from the Calabrian arc reveal arc-orthogonal extension across the subduction zone

The Calabrian Arc subduction-rollback system along the convergent Africa/Eurasia plate boundary is among the most active geological structures in the Mediterranean Sea. However, its seismogenic behaviour is largely unknown, mostly due to the lack of seismological observations. We studied low-to-moderate magnitude earthquakes recorded by the seismic network onshore, integrated by data from a seafloor observatory (NEMO-SN1), to compute a lithospheric velocity model for the western Ionian Sea, and relocate seismic events along major tectonic structures. Spatial changes in the depth distribution of earthquakes highlight a major lithospheric boundary constituted by the Ionian Fault, which separates two sectors where thickness of the seismogenic layer varies over 40 km. This regional tectonic boundary represents the eastern limit of a domain characterized by thinner lithosphere, arc-orthogonal extension, and transtensional tectonic deformation. Occurrence of a few thrust-type earthquakes in the accretionary wedge may suggest a locked subduction interface in a complex tectonic setting, which involves the interplay between arc-orthogonal extension and plate convergence. We finally note that distribution of earthquakes and associated extensional deformation in the Messina Straits region could be explained by right-lateral displacement along the Ionian Fault. This observation could shed new light on proposed mechanisms for the 1908 Messina earthquake.

Seismic velocity structure at the southern termination of the 2016 Kumamoto Earthquake rupture, Japan

The earthquake rupture termination mechanism and size of the ruptured area are crucial parameters for earthquake magnitude estimations and seismic hazard assessments. The 2016 Mw 7.0 Kumamoto Earthquake, central Kyushu, Japan, ruptured a 34-km-long area along previously recognized active faults, eastern part of the Futagawa fault zone and northernmost part of the Hinagu fault zone. Many researchers have suggested that a magma chamber under Aso Volcano terminated the eastward rupture. However, the termination mechanism of the southward rupture has remained unclear. Here, we conduct a local seismic tomographic inversion using a dense temporary seismic network to detail the seismic velocity structure around the southern termination of the rupture. The compressional-wave velocity (Vp) results and compressional- to shear-wave velocity (Vp/Vs) structure indicate several E–W- and ENE–WSW-trending zonal anomalies in the upper to middle crust. These zonal anomalies may reflect regional geological structures that follow the same trends as the Oita–Kumamoto Tectonic Line and Usuki–Yatsushiro Tectonic Line. While the 2016 Kumamoto Earthquake rupture mainly propagated through a low-Vp/Vs area (1.62–1.74) along the Hinagu fault zone, the southern termination of the earthquake at the focal depth of the mainshock is adjacent to a 3-km-diameter high-Vp/Vs body. There is a rapid 5-km step in the depth of the seismogenic layer across the E–W-trending velocity boundary between the low- and high-Vp/Vs areas that corresponds well with the Rokkoku Tectonic Line; this geological boundary is the likely cause of the dislocation of the seismogenic layer because it is intruded by serpentinite veins. A possible factor in the southern rupture termination of the 2016 Kumamoto Earthquake is the existence of a high-Vp/Vs body in the direction of southern rupture propagation. The provided details of this inhomogeneous barrier, which are inferred from the seismic velocity structures, may improve future seismic hazard assessments for a complex fault system composed of multiple segments.

Lithological and structural control on the seismicity distribution in Central Italy

&lt;p&gt;In the seismically active region of Central Italy, national (permanent) and local (not-permanent) seismic networks provided very accurate location of the seismicity recorded during the major seismic sequences occurred in the last 25 years (e.g. 1997-1998; 2009; 2016-2017), as well as of the background seismicity registered in the intervening periods.&amp;#160; In the same region, a network of seismic reflection profiles, originally acquired for oil exploration purposes, is also available, effectively imaging the geological structure at depth, to be compared with the seismicity distribution.&amp;#160;&lt;/p&gt;&lt;p&gt;This comparison reveals that, if the position of the brittle/ductile transition exerts a role at regional scale for the occurrence of crustal seismicity, at a more local scale the depth and thickness of the seismogenic layer is mostly controlled by the contrasting rheological properties of the different lithological groups involved in the upper crust.&amp;#160;&lt;/p&gt;&lt;p&gt;The upper crust stratigraphy, including the sedimentary cover and the uppermost part of the basement, consists of alternated strong (rigid, e.g. carbonates and dolostones) end weak (not-rigid, e.g. shales, sandstones, and phyllites) layers. This mechanically complex multilayer is involved in a belt of imbricated thrusts (Late Miocene-Early Pliocene), displaced by subsequent extensional (normal) faults (Late Pliocene-present), responsible for the observed regional seismicity. The top of the basement s.l. (composed of clastic sedimentary and slightly metamorphosed rocks) is involved in major thrusts. &amp;#160;For these different lithological units, combined field and lab studies of fault rock properties have documented localized and potentially unstable deformation occurring in granular mineral phases (carbonates) and distributed and stable slip within phyllosilicate-rich shear zones (shales and phyllites).&lt;/p&gt;&lt;p&gt;By comparing the geological structure with the seismicity distribution, we observed that:&lt;/p&gt;&lt;p&gt;-&amp;#160;&amp;#160;&amp;#160;&amp;#160;&amp;#160;The seismicity cut-off (i.e. the bottom of the seismogenic layer) is structurally (not thermally) controlled, and grossly corresponds to the top basement; the upper boundary of the seismogenic layer corresponds to the top of carbonates.&lt;/p&gt;&lt;p&gt;-&amp;#160;&amp;#160;&amp;#160;&amp;#160;&amp;#160;Most seismicity occurs within the rigid layers (carbonates and evaporites), and do not penetrate the turbidites and basements rocks.&lt;/p&gt;&lt;p&gt;-&amp;#160;&amp;#160;&amp;#160;&amp;#160;&amp;#160; Close to the axial region of the mountain range, where the larger amount of shortening is observed, the presence thrust sheets from the previous compressional phase, significantly affect the seismicity distribution and propagation.&lt;/p&gt;&lt;p&gt;-&amp;#160;&amp;#160;&amp;#160;&amp;#160;&amp;#160;Major east-dipping extensional detachments, recognized in the seismic profiles, are also marked by distinctive seismicity alignments.&lt;/p&gt;

Monitoring aseismic creep trends in the İsmetpaşa and Destek segments throughout the North Anatolian Fault (NAF) with a large-scale GPS network

The North Anatolian Fault Zone (NAFZ) is an intersection area between the Anatolian and Eurasian plates. The Arabian Plate, which squeezes the Anatolian Plate from the south between the Eurasian Plate and itself, is also responsible for this formation. This tectonic motion causes the Anatolian Plate to move westwards with almost a 20 mm yr−1 velocity, which has caused destructive earthquakes in history. Block boundaries that form the faults are generally locked to the bottom of the seismogenic layer because of the friction between blocks and are responsible for these discharges. However, there are also some unique events observed around the world, which may cause partially or fully free-slipping faults. This phenomenon is called “aseismic creep” and may occur through the entire seismogenic zone or at least to some depths. Additionally, it is a rare event in the world located in two reported segments along the North Anatolian Fault (NAF), which are İsmetpaşa and Destek. In this study, we established GPS networks covering those segments and made three campaigns between 2014 and 2016. Considering the long-term geodetic movements of the blocks (Anatolian and Eurasian plates), surface velocities and fault parameters are calculated. The results of the model indicate that aseismic creep still continues with rates of 13.2±3.3 mm yr−1 at İsmetpaşa and 9.6±3.1 mm yr−1 at Destek. Additionally, aseismic creep behavior is limited to some depths and decays linearly to the bottom of the seismogenic layer at both segments. This study suggests that this aseismic creep behavior will not prevent medium- to large-scale earthquakes in the long term.

Moment–Area Scaling Relationship Assuming Constant Stress Drop for Crustal Earthquakes

ABSTRACT For crustal earthquakes, the scaling relationship between the seismic moment M0 and rupture area S varies with the size of the earthquake, due to the limited thickness of the seismogenic layer. In those M0–S scaling relations, in most cases, the calculated static stress drop is altered with the size of earthquake, although the change depends on the assumed fault model. However, it is not clear whether the dependence of the stress drop on M0 is physically reasonable. In this study, the scaling relation between M0 and S, which assumes a constant stress drop over a wide M0 range, is discussed based on the analytical stress drop formula of a rectangular strike-slip fault. In the proposed relation, M0 is proportional to S3/2 for small and medium faults and to S1 for long faults. In addition, the relation between M0 and S varies in the intermediate range, depending on the aspect ratio. The scaling relation showed good agreement with past event data when the saturated rupture width was set to around 15–20 km and the stress drop was set to about 3 MPa.

A safer future with clues from earthquakes past

Understanding the short- and long-term mechanical behaviour of the lower crust is of fundamental importance when trying to understand the earthquake cycle and related hazard along active fault zones. In some regions some 20% of intracontinental earthquakes of magnitude > 5 nucleates in the lower crust at depth of 30-40 km. For example, a significant proportion of seismicity in the Himalaya, as well as aftershocks associated with the destructive 2001 Bhuj earthquake in India, nucleated in the granulitic lower crust of the Indian shield. Earthquakes in the continental interiors are often devastating and, over the past century, have killed significantly more people than earthquakes that occurred at plate boundaries. Thus, a thorough understanding of the earthquake cycle in intracontinental settings is essential. This requires knowledge of the mechanical behaviour and of the strength (by which Earth scientists commonly mean the maximum stress that rocks can sustain before deforming) of the lower crust. The most common conceptual model of the strength of the continental crust predicts a strong, seismogenic brittle upper crust (where the base of the seismogenic layer is typically considered to be at depth of 10-15 km), and a weak, viscous, aseismic lower crust. This model has been recently questioned by the finding that the lower crust can be seismic and, therefore, mechanically strong. The question arises, how thick is the seismogenic layer in the crust? Answering this question is crucial to determine the potential hazard caused by large earthquakes, which are also generally the deepest.<br/> Our limited knowledge of the mechanical behaviour of the lower crust is largely due to the lower crust itself being poorly accessible for direct geological observations, so that most of our knowledge derives from indirect geophysical measurements (like the distribution of earthquakes). There are only a few well-exposed large sections of exhumed continental lower crust in the world. One of these is located in the Lofoten islands (northern Norway), which were exhumed from their original deep crustal position during the opening of the North Atlantic Ocean.<br/> We propose an integrated, multi-disciplinary study of a network of brittle-viscous shear zones (i.e. zones of localized intense deformation of geological materials) from Lofoten, which records episodic rapid slip events (earthquakes) alternating with long-lasting aseismic creep. The study will link structural geology (analysis of geological faults and shear zones), petrology (analysis of the composition and textures of rocks), geochemistry (detailed chemical characterization of rocks and minerals) and experimental rock deformation (to reproduce in the lab under controlled conditions the deformation processes operative in the deep Earth's crust). This integrated dataset will provide a novel, clear picture of the mechanical behaviour of the continental lower crust during the earthquake cycle. Our direct geological and experimental observations will be tested against geophysical observations of currently active seismic deformation. The cumulative results of the projects will shed light on the currently poorly constrained mechanical behaviour of the lower crust during the earthquake cycle, and therefore on the sequence of inter-seismic slip (the period of slow accumulation of elastic deformation along a fault), co-seismic slip (the sudden rupture along a fault that is the earthquake) and post-seismic slip (the immediate period after an earthquake when the crust and the fault adjust to the modified state of crustal stress caused by an earthquake). This will greatly extend and complement existing efforts by the scientific community to understand and interpret the mechanical behaviour of rocks during the earthquake cycle recorded in the lower crust and the related hazard, and will provide key input for numerical models of continental dynamics.

Study on the layout of GNSS sites for strike-slip faults

SUMMARY In the profile analysis of faults, the distribution of GNSS sites directly affects the accuracy of the results of slip rate and locking depth. This paper discusses strategies for designing the layout of GNSS stations perpendicular to strike-slip faults in terms of site spacing and the Minimum Effective Distance, which is 20 times the locking depth of the fault. Three layout models are proposed considering the complexity of strike-slip faults: (1) Equal spacing layout, in which many stations are deployed in the far field, only a few are deployed in the near field. (2) Equal deformation layout, in which stations are densely arranged in the near field and sparsely arranged in the far field according to the frequency of deformation curve. (3) Equal slope spacing layout, in which stations are arranged according to the nonlinear degree of the deformation curve, with dense distribution in regions with high nonlinearity and sparse distribution in approximately linear regions. The three models were used to redistribute the sites in the Qiaojia to Dongchuan segment of the Xiaojiang fault profile, and their performances were compared with that of the current sites distribution of the segment. The results showed that model 1 is optimal for fitting the accuracy of slip rate and model 3 is optimal for the accuracy of locking depth. Overall, model 3 appears to be the best choice, considering that the accuracy of the locking depth is more difficult to control. One of the main purposes of deployment is to identify the seismogenic depth of the fault. With the locking depth of the fault gradually approaching the depth of the seismogenic layer during an interseismic period, the accuracy of observations of sites deployed at a preset value of historical seismogenic depth of the fault would improve.