Earthquake Science Center Seminars

The release in 1965 of Paramount's science fiction classic, Crack in the World, challenged scientists and engineers to think about the possibility that a giant dike could fracture Earth's brittle crust and encircle the planet. The movie also offered a solution to the hydrocarbon energy crisis and global warming decades before anyone thought these were serious problems.

In this seminar I will replay key cuts from this sci-fi movie thriller and show how geology, volcanology, elasticity theory and fracture mechanics help one to separate fact from fantasy. I will use data reviewed in Townsend et al., 2017, Tectonophysics to argue for a third school of thought regarding mechanical models for dikes. Also, I will use model results published in Pollard & Townsend, 2018, JSG to highlight the role of gravity and a stratified crustal density to stabilize giant dikes, even if the fracture toughness of rock is zero. Finally, I will explore interesting instabilities for fluid-filled fractures predicted by the model results.

U.S. oil fields have been producing much more water than oil for decades. However, injection-induced seismicity has markedly increased with the recent unconventional oil revolution. At the basin-scale, injection volumes have doubled across all plays since 2009, with a key feature being that much of recent injection has occurred into non-producing horizons. Here, I'll discuss a few recent projects addressing injection-induced seismicity at multiple scales using statistical and physics-based approaches. Finally, I'll show how linking hydrogeology and geomechanics in stochastic optimization models can aid future mitigation with targeted injection reduction efforts.

Beginning in August 2016, a series of moderate to large earthquakes (M6.0 August 24, M5.9 October 26, M6.5 October 30) struck the central Apennines producing severe damage in several small towns including Amatrice, Norcia and Visso. The earthquakes resulted in almost 300 casualties and left more than 40,000 homeless. This earthquake sequence is located between the 1997M.6.0 Umbria Marche sequence to the north and the 2009 M6.1 L'Aquila earthquake to the south, defining a continuous 140 km-long NW-striking zone of normal faulting. Large complexity characterizes the sequence in terms of rupture geometry and slip distribution both at depth and at the surface. The seminar includes a review of the present knowledge on the seismic sequence.

Recent advances in more than one dimensional anelastic wave theory predict characteristics of seismic waves in an anelastic Earth that are not predicted by elastic models.

Anelastic solutions of fundamental problems reveal:

1) P and S body waves refracted at anelastic boundaries with wave speeds and attenuation coefficients that vary with contrasts in intrinsic material absorption and angle of incidence,

2) Anelastic seismic rays distinct from those predicted by elastic solutions with measurable differences in ray-path locations, distances, travel times, and amplitude attenuation in the near- surface, crust, and mantle.

3) New insights for interpretation of seismic waves to infer anelastic material characteristics of the Earth's interior.

Continuous seismic monitoring can provide early warning of potentially

hazardous conditions developing in the subsurface, as well as useful information

on the flow of natural resources such as water and hydrocarbons. However, the

cost of continuous monitoring is a significant obstacle to its widespread

application. Distributed Acoustic Sensors (DAS) DAS arrays have the potential of

enabling cost-effective continuous and spatially dense seismic monitoring over

large areas. By leveraging the preexisting fiber infrastructure developed for

telecommunication purposes, we could build large seismic arrays under densely

populated cities around the world in a cost-effective way.

Starting the beginning of September 2016, we have been continuously recording

seismic data as sensed by a fiber-optic cable placed under Stanford University

campus and measured by an OptaSense DAS system. Our "Stanford Fiber

Seismic Observatory (S-FSO)" experiment aims at studying the feasibility of

using a DAS array that record data using a fiber-optic cable "free-floating" in a

PVC conduit buried in the ground. Coupling between the fiber cable and the

surrounding rocks relies exclusively on gravity and friction. Our array has

recorded hundreds of seismic events. Preliminary analysis of these events shows

that the recorded signal is highly repeatable and correlates well wit the data

recorded by a broadband seismometer located on campus. We are applying

ambient (mostly anthropogenic) noise interferometry to estimate and monitor

subsurface properties under the array. Data analysis has also prompted us to

develop new theoretical models to better interpret the recorded signal and the

interferometric data.

Earthquake-induced landslides represent a significant seismic hazard, as evidenced by recent earthquakes in Kaikoura, New Zealand and Gorkha, Nepal, and proper planning/mitigation requires accurate evaluation of the potential for seismic landslides. Engineers often tackle this problem through a detailed evaluation of individual slopes and more recently have introduced performance-based engineering (PBE) concepts into the analysis. Recognizing the compounding effects of multiple landslides across an area, earth scientists often evaluate seismic landslides at a regional scale. This approach sacrifices details, but provides a broader assessment of the impacts of earthquake induced landslides. This presentation will describe the integration of performance-based engineering concepts into regional-scale seismic landslide assessments. The basic PBE framework for seismic landslides will be introduced along with the modifications required to apply it at a regional scale. The application of the approach for a seismic landslide hazard map will be presented. The use of seismic landslide inventories to validate regional landslide assessments will be discussed, along with advancements in developing seismic landslide inventories using remote sensing techniques. Finally, research needs required to further advance regional seismic landslide assessments will be presented.

Convergent tectonic plate boundaries host elevated seismic hazard and spectacular topographic relief. Detailed characterization of earthquake sources (e.g. faults) provide critical insight into the spatial variability of contemporary seismic hazards, while mapping their long-term (i.e. geologic) history can improve our understanding of how topography is generated and evolves over time.

In the Pacific Northwest of the United States, active crustal deformation (faulting, folding, block rotation) above the Cascadia subduction zone is driven by oblique plate convergence (Juan de Fuca-North America) and clockwise rotation of the overriding North American plate. Upper-plate rotation results in distributed transtension in western Idaho and eastern Oregon, which transitions to distributed transpression in northwest Oregon and throughout much of Washington State and southwest British Columbia (Canada). In Washington, transpressional deformation occurs from the Cascadia fore arc to its back arc, at least 500 km from the trench. Unresolved questions persist about how regional strain is distributed over, and transmitted across, a network of low strain rate structures in the upper plate.

To improve our understanding of structural linkages across the Pacific Northwest, we interrogate the latest Pleistocene to Holocene record of crustal deformation along several faults and folds in the Olympic Mountains, Puget Lowland, and Yakima folds province. We employ paleoseismologic, neotectonic, and marine geophysical techniques. Our observations allow us to model the timing of surface-rupturing Holocene paleo-earthquakes, quantify co-seismic folding and uplift, and estimate fault slip rates. We also examine the tectonic geomorphology of structures to characterize their recent growth and understand how local topographic relief is generated. Collectively, these results will aid seismic hazard analyses related to major population centers and critical infrastructure and improve our understanding of the evolving Pacific Northwest landscape over space and time.

Geotechnical engineers and geophysicists are faced with the problem of characterizing geological materials for site-specific design in the built environment. When the design or analysis includes determining the dynamic response of important infrastructure and critical facilities due to dynamic loading such as earthquake shaking, seismically-based measurements in the field are essential for direct characterization of material stiffnesses, particularly shear wave velocities. These measurement techniques were generally adapted from exploration geophysics, required boreholes, and involved small-strain body waves. Today, field seismic measurements are widely used and are evolving because: (1) the measurements have a strong theoretical basis, (2) they can be performed in the field and laboratory, thus forming an important link between these measurements, and (3) in recent developments in the field testing involving surface waves, they are noninvasive which makes them very cost effective. Active-source, surface-wave methods are used today over depths ranging from a few meters to more than 500 m, and passive surface-wave methods are used for even greater depths. Examples of shear-wave velocity (VS) profiles evaluated using Rayleigh-type surface waves are presented. The active-source, Spectral-Analysis-of-Surface-Waves (SASW) method is used to illustrate the measurements. VS profiles have been performed in materials ranging from uncemented soil to unweather rock. Some interesting observations are: (1) the coefficients of variation (COVs) in the VS profiles are generally less than 0.15 over sites with surface areas of 50 km2 or more as long as material types are not laterally mixed, (2) the stiffness of rock near the ground surface is generally overestimated, and (3) when values of VS measured in the field and laboratory are compared, biases in VS at soil sites versus rock sites exhibit opposite trends.

The importance of the damage zone in the faulting and earthquake process is widely recognized, but our understanding of how damage zones are created, what their properties are, and how they feed back into the seismic cycle, is remarkably poorly known. Firstly, damaged rocks have reduced elastic moduli, cohesion and yield strength, which can cause attenuation and potentially non-linear wave propagation effects during ruptures. Secondly, damaged fault rocks are generally more permeable than intact rocks, and hence play a key role in the migration of fluids in and around fault zones over the seismic cycle. Finally, the dynamic generation of damage as the earthquake propagates can itself influence the dynamics of rupture propagation, by increasing the amount of energy dissipation, decreasing the rupture velocity, modifying the size of the earthquake, changing the efficiency of weakening mechanisms such as thermal pressurisation of pore fluids, and even generating seismic waves itself . All of these effects imply that a feedback exists between the damage imparted immediately after rupture propagation, at the early stages of fault slip, and the effects of that damage on subsequent ruptures dynamics.

In recent years, much debate has been sparked by the identification of so-called 'pulverized rocks' described on various crustal-scale faults, a type of intensely damaged fault rock which has undergone minimal shear strain, and the occurrence of which has been linked to damage induced by transient high strain-rate stress perturbations during earthquake rupture. Damage induced by such transient stresses, whether compressional or tensional, likely constitute heterogeneous modulations of the remote stresses that will impart significant changes on the strength, elastic and fluid flow properties of a fault zone immediately after rupture propagation, at the early stage of fault slip. In this contribution, we will demonstrate laboratory and field examples of two dynamic mechanisms that have been proposed for the generation of pulverized rocks; (i) compressive loading by high-frequency stress pulses due to the radiation of seismic waves and (ii) explosive dilation in tension in rocks containing pressurized pore fluids.