Univ. of Houston
It is often suggested that Paleo-Pacific plate subduction dominated the Jurassic-Cretaceous tectonics of East Asia, creating a 6000-km-long arc from Far East Russia to Borneo. Nevertheless, the nature of the Paleo-Pacific are controversial, because it had been entirely consumed. I use seismic tomography models to map the subducted slabs in 3-D and tie them with geologic records. Three slabs are identified in the lower mantle: Izanagi, Hunan, and Taipei. The Izanagi slab is a N35E-trending, 2700-km wide slab wall, located beneath Far East Russia and Northeast China. To the south, the Hunan slab is separated by a gap from the Izanagi. It also dips vertically. Projected onto surface, this slab ranges from the Taihang Mountains in North China, through South China, to northern Vietnam. The Taipei slab dips moderately to the south, and strikes ENE at shallow depths and ESE at deeper depths. It lies under the Taiwan Island and the Philippine Sea with a minimal width of 2200 km. The so-called Paleo-Pacific consists of at least three plates with different subduction history. Their geometry and spatial relationship provide independent evidence and constraints for East Asian tectonics, including: 1) extensive Jurassic-Cretaceous magmatism along the entire margin, 2) a giant flat-slab subduction event in North and Northeast China, 3) subduction cessation in SE China in early Late Cretaceous, 4) existence of a slab tear followed by ridge subduction, 5) rotational, retreating trench during the Taipei slab subduction, and 6) exotic terrane accretion.
Utah State University
Friction-generated heat exerts a fundamental control on fault strength and fault rock rheological evolution during the seismic cycle. Hematite fault mirrors in exhumed fault zones record these transient thermal and mechanical processes during earthquakes. Novel nanoscale textural and chemical tools can document evidence for dynamic weakening and potential re-strengthening of fault surface hematite. Hematite is also amenable to (U-Th)/He thermochronometry and the kinetics of this system responds to short-duration, high-temperature thermal pulses from flash heating of asperities to serve as paleotemperature proxy. Accurate interpretation of hematite (U-Th)/He data requires hematite textural characterization, grain-size (closure temperature) distribution data, and constraints on ambient thermal conditions during and after hematite formation from apatite He thermochronometry. This talk explores innovative applications of nanotextural and nanochemical data and hematite (U-Th)/He thermochronometry to recover transient heat signatures in fault rocks, with implications for dynamic weakening and re-strengthening mechanisms during seismic slip.
Caltech (previously UCLA)
Landslides are a major hazard as well as a primary driver of erosion and geochemical cycles in steep mountainous landscapes. Understanding what controls landslide sizes is central to evaluating landslide hazards and to unraveling tectonics-climate-erosion interactions. In this work, we investigate the influence of topographically induced fracture patterns on bedrock landslide sizes at the eastern margin of the Tibetan Plateau. To do this, we model the subsurface stress fields resulted from the interaction between regional tectonic stress and topographic perturbation. Then, we predict the spatial patterns of fracture-rich zone and compare them with the observed sizes of ~ 1000 bedrock landslides. We use two landslides inventories including earthquake-induced landslides from the 2008 Wenchuan earthquake and climate-induced landslides from before the Wenchuan event. We show that the extent of fracture-rich zones likely limit the sizes of large bedrock landslides. Overall, our observations provide new insights into the controls of landslide dimensions, which can provide critical information on landslide hazard prediction and mitigation.
Luca Dal Zilio,
Assessing the magnitude and return period of large earthquakes along the Himalayan arc is both a major societal concern and scientific challenge. Therefore, efforts to establish a better understanding of the physics governing fault and earthquake dynamics are of utmost importance. In this talk, I will discuss recent progress we have made towards that goal. In the first part, I present physics-based numerical models, which are used to constrain the geometry of the Main Himalayan Thrust (MHT) and to explore the conditions that could explain the bimodal seismicity (Mw ≤ 7.8 vs. Mw 8+) of the large Himalayan earthquakes. These models establish the dependence of earthquake rupture patterns on fault frictional properties and non-planar geometry of the MHT due to variations in strength excess. In the second part, I propose the first probabilistic estimates of interseismic coupling along the MHT based on a recent compilation of geodetic data. Using a fully Bayesian approach, I evaluate the population of plausible coupling models given geodetic data and forward problem uncertainties. I will show how the spatial variability in coupling and complexity in earthquake history suggest lateral segmentation in the collisional structure, which are related to inherited tectonic structures from the India-Eurasia collision. The Himalayan megathrust appears to be paved with low- and high-coupling patches and the resulting pattern seems to have a profound influence on its long-term seismic behaviour, as well as on individual earthquakes.
University of Grenoble
Subduction zones host the largest earthquakes and associated tsunamis. Several subduction earthquakes have been preceded by a precursory activity, namely an increase of the seismicity rate before the mainshock. In this seminar I will discuss examples of how this accelerated seismicity can be associated with other deformation or seismological transient signals at various time and space scales: from localized preseismic slip on the subduction interface before the earthquake, to intriguing long-term interactions between deep seismicity and interface seismicity associated with slow decoupling.
UC Santa Cruz
Understanding processes of slow frictional slip in landslides is key to landslide hazard prediction and mitigation. Additionally, because slow landslides in California are particularly common within the Franciscan Mélange, an exhumed subduction zone accretionary complex, they may provide a window into slip processes at work deep within subduction zones. In this talk, I will use hydrological, deformational, and shallow geophysical monitoring of a large earthflow in California’s Diablo Range to explore 1) what triggers the onset of seasonal slip in California, 2) why earthflows are apparently so close to frictional failure at all times, and 3) whether the prevailing model for slow frictional slip in landslides, dilatant strengthening, is consistent with our detailed observations.
Korean Peninsula is an intraplate region located ∼800 km away from the Ryukyu and Nankai trough, southwestern Japan. According to an earthquake catalogue published by the Korea Meteorological Administration (KMA), only 10 earthquakes with M≥5.0 have been instrumentally recorded since 1978. The largest among them, a M5.5 earthquake in Gyeongju, South Korea on 2016 September 12, produced strong coseismic ground shaking, which was sufficient to be felt throughout South Korea. Less than a year later, M5.4 earthquake occurred near the enhanced geothermal system (EGS) site in Pohang (~30 km northeast from the Gyeongju earthquake), and this M5.4 earthquake is the second largest and the most damaging event in South Korea, and furthermore is the first reported, largest induced earthquake, occurred at any EGS site worldwide. The M5.5 Gyeongju earthquake happened in the Gyeongsang Basin in southeastern Korean Peninsula, where there are several systematic faults with surface expression such as the Yangsan Fault Systems. The 2016 Gyeongju earthquake sequence started from a M5.1 event that ruptured 50 min before the M5.8 event, and thousands of earthquakes, including M4.3 aftershock on September 19, occurred in the sequence. The source mechanism of the foreshock, mainshock, and aftershock events (M5.1, 5.5, and 4.3, respectively) shows strike-slip motion on the fault system at a depth range of 14–16 km, based on the regional waveform inversion (Kim et al., 2017). The hypocenter of the M5.4 Pohang earthquake was at ~4–5 km depth and ~10 km east of the Yangsan fault. The space-time variation of seismicity before this earthquake is well correlated with the history of stimulation activities involving fluid injection and withdrawal (Kim et al., 2018). As the seismicity in Pohang area is very low before the EGS operation, the proximity of the focal depth to the well tip and temporal correlation between seismicity and hydraulic stimulation support the notion that geothermal plant activities may cause the M5.4 Pohang earthquake. Using the fluid injection history, we investigate the dependence of stress change with respect to poroelastic and injection parameters and compare seismicity against the spatiotemporal evolution of Coulomb stress change. In this study, we determine the locations of the mainshock as well as 311 smaller earthquakes, including nine precedent earthquakes and 302 aftershocks from December 2016 to February 2018, and provide tight constraint linking the stress change to the earthquake location.
We study temporal changes of seismic velocities associated with the June 2016 Mw 5.2 Borrego Springs earthquake in the San Jacinto fault zone, using 9 component Green’s function estimates reconstructed from daily cross correlations of ambient noise. The analyzed data are recorded by stations in two dense linear arrays crossing the fault surface trace ~3 km northwest (DW) and southeast (JF) of the event epicenter. The two arrays have 9-12 stations each with instrument spacing of 25-100 m. Relative velocity changes (δv/v) are estimated from arrival time changes in the daily correlation coda waveforms compared to a reference stack. The obtained array-average δv/v time series exhibit seasonal variations, linear trends and changes associated with the Borrego Springs event. The earthquake signal is characterized by rapid co-seismic velocity drop followed by a gradual recovery. This is observed consistently using both time- and frequency-domain δv/v analysis methods on data of different components in various frequency bands at both arrays. After removal of the seasonal and linear components, the velocity drops estimated with a daily resolution from the vertical-vertical correlation data are around 6%, 0.3%, and 0.2% at the DW array, and 1%, 0.3%, and 0.2% at the JF array, for 0.1-0.4 Hz, 0.5-2 Hz, and 1-4 Hz, respectively. Normalizing the co-seismic changes by corresponding amplitudes of seasonal variations helps suppressing site-specific responses. The larger changes at lower frequencies imply that the variations are not limited to the subsurface material. A decreasing co-seismic velocity reduction with coda wave lapse time indicates larger co-seismic structural perturbations within the fault zone compared to the surrounding host rock. The observed larger changes at the DW array compared to the changes at the JF array may reflect larger damage produced by the northwestward rupture directivity of the Borrego Springs earthquake.
I present new seismic images of the key structural elements of the upper mantle beneath North America. These images are significantly different from all previous studies. Unlike earlier work, we successfully image the cratonic lithosphere-asthenosphere boundary (LAB), and the 250-km-deep Lehmann discontinuity. We identify a pronounced, if enigmatic, low velocity layer (LVL) above the mantle transition zone (410-discontinuity). These results have important implications for the evolution of the upper mantle.
Erik Weidner, Han Bao,
Han Bao: The speed at which an earthquake rupture propagates affects its energy balance and ground shaking impact. Supershear earthquakes (faster than shear-wave speed Vs) often start at sub-shear speed and later run faster than Eshelby’s speed (√2Vs). Here we present robust evidence of early and persistent supershear rupture at sub-Eshelby speed of the 2018 Mw 7.5 Palu, Indonesia earthquake. Slowness-enhanced back-projection of teleseismic data provides a sharp image of the rupture process, along a path consistent with the surface rupture trace inferred by sub-pixel correlation of interferometric synthetic-aperture radar images. The rupture propagated at a sustained velocity of 4.1 km/s from its initiation to its end, despite large fault bends. The persistent supershear speed is further validated by seismological evidence of far-field Rayleigh Mach waves. The unusual features of this earthquake may reveal connections between rupture dynamics and fault structure. Early supershear transition can be promoted by fault roughness near the hypocenter. Steady rupture propagation at a speed unexpected in homogeneous media could result from the presence of a damaged fault zone.