CNRS Ecole Normale Superieure de Lyon
|Rheology of serpentines, seismicity and mass transfer in subduction zone|
Dr. Gerrie van Aswegen
|Three faults, five earthquakes in Welkom|
|Evidence of dike feeding system at Stromboli volcano|
Hermann M. Fritz
Georgia Institute of Technology
|Tsunami generation by landslides and extreme runup heights|
Prof. Miles B. Rubin
Israel Institute of Technology
|Physical reasons for abandoning plastic deformation measures in plasticity and viscoplasticity theory|
John G. Anderson
University of Nevada, Reno
|Source and Site Characteristics of Earthquakes that have Caused Exceptional Ground Accelerations and Velocities|
|The origin of deep ocean microseisms and implications for noise tomotgraphy studies|
Bristol University; ERI visiting professor
|Mantle structure: The message from scattered waves|
Dr. Laura Wallace
GNS Science, New Zealand
|Characterizing the seismogenic zone of a major plate boundary subduction thrust: the Hikurangi Margin, New Zealand|
|Grain size evolution in the mantle|
Serpentinites have a lower density and lower viscosity than "dry" ultramafic rocks and it was proposed, based on numerical simulations, that they play a major role in mantle-slab decoupling, and in downward (sink) or upward (exhumation) motion of eclogites and ultra-high pressure (UHP) rocks in subduction zones. Rheological data on antigorite, the stable variety of serpentine in subduction zones, were obtained over a P-T range of 1-4 GPa and 200-500 /deg C that cover most of its stability field. The experiments were carried out in a D-DIA apparatus installed at GSECARS on the 13-BM-D line of APS. The determined stress-strain curves were fitted to a power-law equation including both temperature and pressure dependence. The results confirm that serpentinites acts as a weak layer that allows significant mass transfer along the "serpentinized channel" and dynamic processes such as mantle slab decoupling, and mantle wedge convection. Regardless of the temperature, serpentinized mantle at the slab surface has a low viscosity that allows localizing the deformation and impeding stress build-up. It will limit the downdip propagation of large earthquakes, and allow viscous relaxation as an origin of post-seismic deformations and slow earthquakes. Models of growth and transport of a serpentinized channel using available kinetic and present rheological data explain high exhumation rates of eclogites and limited thickness of the channel at great depths (? 50 km), and slower exhumation in a thick hydrated mantle corner at shallower depths.
Such channels may be difficult to detect from sismic tomography or using guided waves because of their small thickness (<2-3 km).
The Dagbreek fault yielded the first major mining-induced earthquake in Welkom in 1976. It failed again in 1988 and the last time in 1999 - the last one being the "Matjhabeng" earthquake. The latter is described in terms of the seismic gap remaining after the first two earthquakes and also in terms of some precursory activity and the possible triggering affect of a mine fire at the time. The Brand fault failed in 1989 - a textbook example of hoe to make an earthquake. The Stuirmanspan earthquake in 1990 showed the practical application of apparent stress as a spatial indicator of the future earthquake source.
Stromboli volcano has erupted on February 27, 2007, after an intense strombolian activity lasted a couple of months. The eruption was characterized by a series of rapid evolving phenomena, like the propagation of the effusive fracture along the crater rim, an unusually large effusive flux (>10 m3/s) and the collapse of the crater system.
We have monitored this eruptive phenomena with an integrated network of multiparameters instruments: broad-band seismometers, infrasonic array, thermal cameras and bore-hole tiltmeters. All these information allowed to draw a quite clear picture of the dynamics of a shallow dike feeding system before and during the onset of the eruption.
高分子ゲルのようなやわらかい物質を平滑基板上ですべらせると， 時空間的に不均一なスティック−スリップ運動を生じることが知られています．最近私たちのグループでは，東急ハンズで売られている粘着性ゲルシートの一端を掴んでガラス基板上で引っ張ると， ある引張速度（すべり速度）を境にスティック−スリップ運動が規則的な振る舞いから不規則的な振る舞いへと転移することを見出しました．また，１回のスリップイベントにおける摩擦力降下量の頻度分布が，低引張速度でべき乗分布（地震学におけるグーテンベルグ−リヒター則）に従い，べき指数（b値）が 引張速度とともに変化することが分かりました．
Tsunamis are commonly associated with submarine earthquakes, such as the 2004 Indian Ocean tsunami. Tectonic tsunamis are limited in wave height by the seafloor displacement. For some earthquakes, such as the 1998 Papua New Guinea and 2006 Java tsunamis, it has been proposed that the large tsunamis were triggered by massive failure of the sea floor in the form of giant submarine landslides. Landslide and volcanic eruption generated tsunamis are typically regionally confined but account for all known localized runup heights exceeding 100 m in the past century: Lituya Bay, Alaska, Vajont Reservoir, Italy and Spirit Lake, Mount St. Helens. Other localized runup heights exceeding 50m have been recorded on Shimabara peninsula and several fjords in Norway and Chile among others. The largest mega-tsunami runup dates back to 10 July 1958, when an earthquake of Mw 8.3 at the Fairweather Fault triggered a rockslide into Lituya Bay. The rockslide impact generated a giant tsunami at the head of Lituya Bay resulting in an unprecedented tsunami runup of 524 m marked by forest trim line and erosion down to bedrock. The mega-tsunami runup is studied with a hybrid modeling approach applying both physical and numerical models of slide processes of deformable bodies into a U-shaped trench similar to the geometry found at Lituya Bay. Tsunami generation by landslides were also studied in the three dimensional NEES (Network for Earthquake Engineering Simulation) tsunami wave basin at OSU (Oregon State University) based on the generalized Froude similarity.
Constitutive equations, which characterize the response of a material to future loadings, must depend on state variables that, in principle, can be measured without any prior knowledge of the past history of deformation of the material. This notion of state is consistent with that proposed by Onat and it is consistent with Gilman's comment on physical problems with using total strain as a state variable in plasticity theory. Within the context of this notion of state, elastic strain is a state variable, whereas the total strain and plastic strains are not state variables since they are measured relative an arbitrary reference configuration. Alternative constitutive equations which are formulated in terms of elastic deformation measures have been discussed in the literature for finite deformations of elastically isotropic and anisotropic elastic-plastic and elasticviscoplastic materials. These constitutive equations have the physical properties that they are independent of the choice of the reference configuration, and they do not utilize any measures of total deformation or plastic deformation. The main objectives of this paper are to discuss physical reasons for abandoning plastic deformation measures in plasticity and viscoplasticity theory and to present an alternative small deformation theory which is formulated in terms of elastic strain. Also, aspects of alternative finite deformation theories are reviewed.
This study investigates the characteristics of the freefield strong-motion records that have yielded the 100 largest peak accelerations and the 100 largest peak velocities available from any of several data sources through July, 2007. The peak is defined as the maximum zero-to-peak amplitude of the acceleration or velocity vector. This compilation found 35 records with peak acceleration greater than 1g (980 cm/s^2), and 41 records with peak velocity greater than 100 cm/s. The results sample an estimated 150,000 instrument-years of strong-motion recordings. The geometric mean of the two horizontal components of acceleration or velocity, as used in many ground motion prediction equations, is typically 0.76 times the magnitude of this vector peak. Accelerations in the top 100 come from earthquakes as small as magnitude 4.8, while velocities in the top 100 all come from earthquakes with M 5.7. These records are dominated by crustal earthquakes with thrust, oblique-thrust, or strike-slip mechanisms. Normal faulting mechanisms in crustal earthquakes constitute under 5% of the records in the databases searched, and an even smaller percentage of the 100 largest acceleration or velocity records. All NEHRP site categories have contributed exceptional records, in proportions similar to the extent that they are represented in larger databases.
Correlation of ambient seismic noise is now widely used to determine
earth structure. The reason for this is the relative ease of
determining the Green's functions along station to station paths compared to the complexity that an earthquake source introduces,and
because of the ability to do this in absence of earthquake sources. The
technique has been applied to both short and long-range surveys in many parts of the world.
Recently, correlation methods are being applied to the problem of monitoring temporal variations in the subsurface. For this application the technique would appear almost ideal because the source is omnipresent compared to earthquakes. However, the need to look at correlations for relatively short time windows can lead to a violation of the underlying assumption of the technique. That is that the sources need to be distributed randomly off (either) end of the station-station path. If this assumption is not met, the technique estimate can be biased by a favored projection of the Green's function. This will lead to an incorrect travel time estimate and consequently an incorrect velocity estimate.
To enable reliable 4D noise tomography (dominated by microseisms), the source must be characterized in space and time. Most microseisms source studies to date are based on correlations between seismic data, buoys, barometer measurements and wave height models. While these statistical associations do reflect a causal relationship, they do not unambiguously highlight the physical process at work nor the exact generation areas or depths. To do that, physical modeling of wave-wave interactions in the open ocean as well as in coastal regions is required. Moreover, the complete theory of microseisms generation in a compressible ocean must be used. While this can be accomplished in a straightforward manner in the open ocean, it is far more challenging in coastal regions, where the wave reflection characteristics is likely to depend on a number of variables, such as the beach slope, the wave height and tidal conditions.
In this presentation we report explicit calculations of microseisms amplitudes making use of hindcast ocean wave spectra from the North Atlantic Ocean, and comparison of the calculations to seismic data collected at stations in North America, Greenland, and Iceland. We find that a particularly energetic source area stretches from the Labrador Sea to a region south of Iceland, where climatological conditions are conducive to generating oppositely traveling waves of same period, and where the ocean depth corresponds to an "organ-pipe" resonance of the compression waves generated by the opposing wave-wave interaction, as predicted by the theory. It is demonstrated that deep ocean nonlinear wave-wave interactions are sufficiently energetic to account for the observed seismic amplitudes in North America, Greenland and Iceland. Differences between Atlantic and Pacific microseisms are highlighted.
When Francis Birch named the Transition Zone, the deep mantle became a dull place. It was homogeneous material simply becoming denser as pressure increased with depth. No more respect was accorded to it by geochemists than by geophysicists. For geochemists, the deep mantle was simply a dark box in which chemical components were held until needed for delicate flavoring of various sorts of rock cocktails. It deserves more respect. Though it may be dregs, the part of the mantle in contact with the core is rich in seismologically annoying structural detail. This might be written off as an observational quirk due to a mendacious Earth or investigative incompetence, except that more of the lower mantle is grudgingly revealing structure as well. The structural details are fine-scale, at characteristic sizes of around one to one hundred kilometers. The details are emerging from studies of scattered seismic waves. These are unscheduled arrivals in the timetable following an earthquake. They don't arise in a uniform or even a layered Earth. Rather, they originate from the wave field's interactions with sub-wavelength roughness in Earth structure. A lot of data is needed to be sure those arrivals are real and repeatable, but networks of hundreds of seismometers such as the ones in existence in Asia, Europe and North America can provide or have provided the necessary redundancy for confident detection. The results of studies of S-to-P and P-to-P scattering to date show that some lower mantle heterogeneity is associated with present subduction. Some is also found at sites of past subduction, but it is difficult to generalize to all heterogeneity. Scattering strength varies with depth: the shallowest lower mantle is rougher than the deeper parts. The peak scattering strength is around 1600 km deep in the mantle, followed by a slow decline. The roughness clusters, too, with individual groups separated by around 100 km. Individual clusters appear to have particular fabrics that influence their scattering characteristics. Because the km- to 100 km-length scales are present in oceanic plates in their layer thicknesses and plate thickness, these features strongly suggest that the scattered waves emanate from solid material injected into the lower mantle by subduction. They also suggest that the deep mantle is not strongly layered in viscosity or density because scattering strength depth profiles do not change abruptly. A real puzzle is the material identity of the heterogeneity. Seismic wavespeeds must change by more than 4% within a kilometer. Clearly, this is no thermal signal, but compositional differences that extreme in mantle mineralogies require extreme variations in silica or a very broad pressure-dependent phase transition to change properties that significantly. Only about 2% of the lower mantle volume has been explored to date. Much of the mantle away from subduction zones will never be visible. Different methods will be needed to see all of the mantle's structure details, even using scattering.
The Hikurangi subduction margin, New Zealand, has not experienced any significant (> Mw 7.2) subduction interface earthquakes since historical records began about 170 years ago. Geological data in parts of the North Island provide evidence for possible pre-historic great subduction earthquakes. Despite the lack of confirmed historical interface events, recent geodetic and seismological results reveal that a large area of the interface is interseismically coupled, along which stress could be released in great earthquakes. I will overview existing geophysical and geological data that we have used to characterize the seismogenic zone of the Hikurangi subduction interface. Deep interseismic coupling of the southern portion of the Hikurangi interface is well-defined by interpretation of GPS velocities, the locations of slow slip events, and the hypocenters of moderate to large historical earthquakes. Interseismic coupling is shallower on the northern and central portion of the Hikurangi subduction thrust. The spatial extent of the likely seismogenic zone at the Hikurangi margin cannot be easily explained by one or two simple parameters. Instead, a complex interplay between upper and lower plate structure, subducting sediment, thermal effects, regional tectonic stress regime, and fluid pressures probably controls the extent of the subduction thrust's seismogenic zone. I will also discuss some striking similarities between the Hikurangi margin and subduction thrust boundaries in northern and southwest Japan.
Y. Ricard and D. Bercovici, A grained continuum model of damage and