S43B-1879
Seismic Wave Strain, Rotation, and Gradiometry for the 4 March 2008 TAIGER Explosions
Acceleration spatial gradients, horizontal strains and horizontal rotation were computed using strong motion array data from the 4 March 2008 Taiger explosions in northeastern Taiwan, and used in conjunction with the original 3 component acceleration data to perform a gradiometric analysis of the strong ground motion wave train. The analysis yields a complex, frequency-dependent view of the nature of seismic wave propagation over short propagation distances that imply significant lateral velocity changes in structure. Areal strain and rotation about the vertical have equal amplitudes and suggest significant wave scattering within the confines of the river valley where the experiment was performed and/or significant departure from an axisymmetri explosion source. Gradiometry shows that the P wave arrives at the array 35 degrees off-azimuth clockwise from the straight-line path and appears to have been refracted from the northern side of the valley. Large, slowly propagating secondary surface waves initially arrive 45 degrees counter-clockwise from the straight- line path but later arrivals are seen to propagate in all directions, including back towards the explosion source. Frequency-dependent radiation pattern for the three-borehole explosion in comparison to the single-borehole explosion explains the differences in maximum amplitudes between sources seen in the acceleration data. The use of seismic strain and rotation with standard particle motion wave fields at a single location allows for a direct view of seismic wave propagation that illuminates the true nature of the seismogram.
S43B-1880
Waves in Linear Elastic Media with Microrotations: Isotropic Full Cosserat Model
In this study, we discussed solutions for longitudinal and transverse bulk waves, Rayleigh wave and surface transverse wave in a half-space as well as Lamb wave and transverse wave in a thin layer within the framework of the isotropic Cosserat continuum. The medium deformation in this model is described not only by the displacement vector, but also by kinematically independent rotation vector. Problems on bulk wave propagation are rather interesting from the viewpoint of interpretation of new parameters introduced for the dimensionless notation of wave solutions. However, these problems have no prospects from the standpoint of experimental implementation. In contrast to bulk waves, surface waves for the elastic half-space have a considerable experimental potential. These waves can be divided into two groups, one of which corresponds to the well-investigated elliptical wave and the other - to the transverse wave with depth-dependent decay, which does not have any analogy in the classical theory of elasticity. The results obtained in this study can be used for the preparation and interpretation of seismic experiments, which could validate the importance of asymmetric theories of elasticity in earthquake and exploration seismology, and for experimental determination of the material constants of the Cosserat media. Firstly, the dispersion analysis of experimental three-component seismograms allows constructing experimental dispersion curves and comparing them with theoretical ones. Secondly, rotation sensors make it possible to trace a relation between displacement and rotation components according to relations. Thirdly, the presence of a transverse surface wave in the Cosserat continuum can also be the subject of experimental study with the use of sensors imbedded at different depths.
S43B-1881
Instrument correction for 6DOF seismic sensors
Pendulum seismic transducers are also sensitive to rotational motions, and record a mixture of translations
and rotations. Retrieving the "pure" translations requires that the rotations are also recorded independently
at the same "point", and that the nature of the coupling is known. Rotational sensors are now being
increasingly deployed in the filed, and the data on directly recorded rotational motions during earthquakes
are increasing. Recordings from six degrees-of-freedom (6 DOF) sensors (recording three components of
translations and three rotations at a "point") make it possible, at least in principle, to retrieve the "pure"
translations provided that appropriate software is available. This paper will present an algorithm for such a
correction for a generic 6DOF sensor, being developed by the authors, and results using synthetic but
realistic 6 DOF motions. The possibility to retrieve the "pure" translational motions will extend the limits of the
useable information that can be obtained from the traditional high dynamic range and high resolution
translational sensors, e.g. information on complex motions and residual deformation of the ground or of
structures, associated with faulting or with nonlinear response.
http://www.usc.edu/dept/civil_eng/Earthquake_eng/
S43B-1882 INVITED
Measuring both Rotational and Translational Ground-Motions from Explosions and Local Earthquakes in Taiwan
Since rotational motions can ”„contaminate”¦ translational ground-motion measurements due to the induced perturbation of the Earth”¦s gravitational field, we started a program to measure rotational ground motions near Hualien (Taiwan) in December, 2000. However, no useful data were obtained after 3 years, until a rotational sensor of much higher sensitivity was deployed at the HGSD station in eastern Taiwan in December, 2004. Rotational and translational seismograms were obtained from several hundred local earthquakes. As noted by several authors before, we found a linear relationship between peak rotational rate (PRR in mrad/s) and peak ground acceleration (PGA in m/s2) from local earthquakes in Taiwan: PRR = 0.002 + 1.301 PGA, with a correlation coefficient of 0.988. Taking advantage of two large explosions of the TAIGER Active Seismic Experiment, we deployed 13 accelerometers and 8 rotational sensors within 600 m from the N3 shot points and obtained some interesting results, which will be presented by Langston et al. in this Session. In December, 2007, we began an instrument array deployment along the Meishan fault in southwestern Taiwan, where a major earthquake occurred in 1906 with surface rupture of more than 12 km long. The deployed instruments are: (1) a 32-element seismic array in free-field, (2) a 32-element accelerometer array in a building, (3) a six-channel unit with a low-gain broadband seismometer and an accelerometer, and (4) two six-channel units with an accelerometer and an external rotational senor. We have 8 rotational sensors now deployed in Taiwan and seven new rotational sensors are scheduled for deployment soon in a program to assess the effect of ground rotation on traditional measurements of translational strong ground motions.
S43B-1883
Observation and Prediction of Dynamic Ground Strains, Tilts and Torsions Caused by the M6.0 2004 Parkfield, California, Earthquake and Aftershocks Derived From UPSAR Array Observations
The September 28, 2004, Parkfield, California, earthquake (Mw 6.0) and four aftershocks (Mw 4.7 - 5.1) were recorded on 12 accelerograph stations of the UPSAR seismic array, an array of three-component accelerographs occupying an area of about 1 square km located 8.8 km from the San Andreas fault. Peak horizontal acceleration and velocity at UPSAR during the mainshock were 0.45 g and 0.27 m/s, respectively. We determined both time-varying and peak values of ground dilatations, shear strains, torsions, tilts, torsion rates, and tilt rates by applying a time-dependent geodetic analysis to the observed array displacement time series. Array-derived dilatations agree fairly well with point measurements made on high sample rate recordings of the Parkfield-area dilatometers (Johnston et al., 2006). Torsion Fourier amplitude spectra agree well with ground velocity spectra, as expected for propagating plane waves. A simple predictive relation, using predicted peak velocity from the Boore-Atkinson (2007) ground motion prediction relation scaled by a phase velocity of 1 km/s, predicts observed peak Parkfield and Chi-Chi rotations (Huang, 2003) well. However, rotation rates measured during Mw 5 Ito, Japan, events observed on a gyro sensor (Takeo, 1998) are factors of 5 - 60 greater than predicted by our predictive relation. This discrepancy might be caused by a scale-dependence in rotation, with rotations measured over a short baseline exceeding those measured over long baselines. An alternative hypothesis is that events having significant non-double-couple mechanisms, like the Ito events, radiate much stronger rotations than double-couple events. If this is true, then rotational observations might provide an important source of new information for monitoring seismicity in volcanic areas.
S43B-1884
Observations And Modeling Of Rotational Signals In The P-Coda: Constraints On Crustal Scattering
In addition to three classical components (vertical, N-S, and E-W) of ground translations recorded by broadband seismometer, a component of earthquake induced rotational ground motions around the vertical axis is consistently measured by a ring laser sensor located in Wettzell, SE Germany. Significant rotations around the vertical axis in the P coda of tele-seismic signals are either directly visible or can be inferred through the investigation of cross-correlation between transverse component of translation acceleration and ring laser rotation rate. Theoretically, in spherically symmetric isotropic media we should not observe rotational signals around the vertical axis before the onset of SH waves. Possible explanations for the observed rotations in the P coda are: (1) tilt ring laser coupling; (2) anisotropy; (3) topographic scattering; and (4) P-SH scattering in the crust. Here we show that P-SH scattering in the 3D random crust is the main cause of the observations and allow us to constrain crustal scattering properties.
S43B-1885
Rotational Motions of Seismic Surface Waves in a Laterally Heterogeneous Earth: Theory and Application to Data
We present a full ray theory (FRT) method to simulate rotational motions of seismic surface waves in smooth, laterally heterogeneous Earth models. In the ray picture of wave propagation the vertical component of the rotational rate motion of fundamental mode Love waves is obtained by dividing the transverse component of ground acceleration by the Love wave local phase velocity beneath the seismic recording station. We illustrate the method with examples of theoretical calculations of T ~ 40s rotational rate ground motions of fundamental Love waves using the crust model CRUST2.0 combined with the mantle model S20RTS for the M8.1 September 25 2003 Tokachi-oki earthquake, Japan. FRT rotation synthetics match complete calculations using the spectral-element method very well and fit real data reasonably well. Furthermore, we show that the effect of realistic local structures beneath receivers on rotational motions is strong enough to be observable. FRT calculations could potentially help inferring Love wave local dispersion curves and thus estimating the 1-D local shear velocity structure beneath seismic stations from point measurements of rotational rate and acceleration ground motions. We support the theoretical work with applications to data from the ring laser rotation sensor at Wettzell, Germany. We estimate Love-wave dispersion curves by stacking spectral ratios of rotation rate and transverse acceleration.
S43B-1886
Comparison of two Different Methods for Recording Rotational Motions: a Commercial Eentec R1 versus Array Derived Rotation
The application of rotational motion sensors already proofed to give new ways of measuring seismic wave field properties when comparing the recorded data with seismograms of collocated traditional seismometers. However, the data of these rare test cases were produced either using sophisticated ring laser technology or cumbersome seismic array techniques including some restrictive assumption about the wave field. In addition, all applications presented so far were performed in the far field of the seismic source. In this paper we want to test the performance of one of the first medium priced, commercial rotational motions sensor (Eentec R1) by comparing its output with the aforementioned array derived rotational motions. The data sets for testing consist of seismic array and rotational motion measurements which were recorded 1) during a demolition blast of a 50 m high building in the city of Munich (Germany) and 2) during frequent volcanic explosions at the Mt. Yasur volcano (Vanuatu) as well as some simple laboratory experiments. At least in case of Mt. Yasur the measurement site was within the near field of the seismic source. As a first step in the analysis chain, we simply compare the various outputs. As next step, we classify the performance of the two methods for recording rotational motions by comparing derived wave field properties with the result of more classical seismic array analysis. The results of this experiments demonstrate, that when using array technique for estimating rotational motions much effort in site selection, array design and a priori knowledge of subsurface conditions is needed. It becomes also evident that the performance of an array and its estimated quantities depends strongly on the number of deployed seismic stations. Given the uncertainties in both the array- derived measurements and the rotation sensor transfer function, which in the latter case are apparent when taking into account effects of translational motions on the rotational sensor, it is difficult to quantify the accuracy of the rotation sensor data. Thus further extensive laboratory and field testing are needed in order to further exploit the new measurable of rotational motions for standard seismological analysis.
S43B-1887
Development of a Torsional Seismometer for measuring the rotational oscillations of the Earth.
The motivations for the development and characterization of instruments capable of recording the rotations associated with seismic activity and normal mode oscillations were detailed extensively at the 2006 Fall- meeting of the American Geophysical Union in 2006 and in a special workshop at USGS-Menlo Park in 2007. This paper describes the effort and progress we have made in building a new instrument to be used for such measurements. Our prototype has two basic subsystems; a torsional oscillator and an optical lever for angular measurements. The essential idea behind the design maybe briefly stated as follows: A mechanical torsional oscillator with a natural frequency significantly below the lowest normal mode frequencies will couple negligibly to the rotational motions of the earth, even though the housing of the oscillator is firmly fixed to the earth. A sensitive optical lever, fixed to the Earth, observing such a balance can therefore faithfully measure the rotational oscillations of the Earth. The challenges we face in this development are two-fold: (a) the development of a mechanical torsional oscillator with a low enough natural frequency ~10- 3Hz and the fabrication of an optical lever with an angular resolution better than ~10- 6rad·Hz-1/2; (b) to make the instrument robust and field-worthy for the study of near-field strong motions at frequencies higher than ~10-2Hz. The initial implemented design is as follows: the balance bob consists of a circular mirror of diameter ~ 40mm, with its normal in the horizontal plane. The mirror is mounted within an aluminum framework whose moment of inertia may be adjusted as required and also used for capacitive damping of unwanted torsional oscillations. The configuration has a mass of under 50g and a moment of inertia of ~150g·cm2 about the suspension axis. The suspension fiber is made of SS-304 alloy with a cross section of 7μm × 110μm and length ~5cm. The angular frequency of natural oscillations for this initial design is ~3×10-2 rad·s-1, which corresponds to a period of ~200s or a frequency of ~5×10-3Hz, i.e. significantly smaller than the frequencies of interest. The second subsystem is a robust yet sensitive optical lever which consists of a slit illuminated by a high intensity LED (50,000mcd) emitting in a forward cone of angle ~7°. The slit is located at the focal plane of a lens of aperture f=200mm. This optical design ensures that the image quality and the angular displacement of the image due to motions of the mirror are sensibly independent of changes in the temperature of the surroundings. The optical image falls on a position sensitive diode whose positional accuracy is ≈3×10-5 mm· Hz-1/2 which corresponds to an angular displacement of the mirror by ≈7.5×10-8 rad·Hz-1/2 Currently we are working to improve the seismometer by adding a position control system to help us acquire reliable data in the presence of torsional oscillations. We are also shielding the system from noise due to radiometric and convective currents. We will report on the performance of this balance operated in the basement of a building at the university.
S43B-1888
Inferring Near-Receiver Structure From Combined Measurements of Rotational and Translational Ground Motions
We present a novel approach to the solution of structural inverse problems using combined measurements of rotational and translational ground motions. This is based on the definition of an apparent shear velocity, β_a, which is the ratio of the rms displacement velocity and the rms rotation. For a pure S wave in a homogeneous medium the apparent shear velocity is equal to the true shear velocity, β. Measurements of β_a are independent of the magnitude and timing of an event. The sensitivity of β_a with respect to β is confined to a volume surrounding the receiver - in contrast to sensitivities for travel times or amplitudes that are nonzero in a volume around the entire ray path. This suggests that the apparent shear velocity may be used for local tomography with teleseismic data even when the correct magnitude and timing are unknown. Inferring Earth structure from measurements of β_a may therefore be a complement to classical receiver function studies and an alternative to local earthquake tomography in regions with little or no seismicity. In this study we explore how the apparent shear wave speed may be used for the solution of structural inverse problems. Our approach is based on the combination of the adjoint method and ray theory for the computation of sensitivity kernels of β_a with respect to Earth structure. We discuss the morphology of the kernels as a function of frequency, and we illustrate how measurements of the apparent shear wave speed at the surface depend on lateral heterogeneities. Our emphasis is on the solution of synthetic structural inverse problems. For this we use an iterative optimisation scheme that is based on the previously computed sensitivity kernels. We demonstrate the functioning of this method with different receiver geometries and frequency ranges.
S43B-1889
Study of rotational ground motion in the near field region
While the earthquake induced translational wave field has been recorded and studied since the nineteenth century, the rotational motion still nowadays remains poorly observed and investigated. We aim at further understanding the rotational ground motion and its relation to the translational wave field with a special emphasis on the near field, few wavelengths away from the hypocenter, where damage related to rotational motion might need to be considered. A broad picture of the available values of rotational amplitudes and their variability is obtained by gathering most of the published data on strong rotational motion. To obtain a more detailed picture, we perform a large scale 3D numerical study of a strike-slip event in the Grenoble valley, where a combination of topographic, source, and site effects produces a realistic wave field. The size of the synthetic dataset allows us to study the distribution of the rotational and translational peak amplitudes and their dependence on two effects: non-linear soil behaviour and source directivity. Finally, we compare our numerical results in terms of Peak Ground Velocity, PGV, vs. Peak Ground Rotation, PGω, with field data obtained at similar scenarios (e.g. Parkfield) by array techniques to investigate the relation between translational and rotational amplitudes which can be expected in the near-field for shallow medium-sized earthquakes. Furthermore, the spatial variations of PGV/PGω ratio show a trend that seems to be correlated with the velocity structure of the model under study.
S43B-1890
Response of Pendulums to Translational and Rotational Components of Ground Motion
Dynamic response of most seismological instruments and many engineering structures to ground shaking can be represented via response of a pendulum (single-degree-of-freedom oscillator). Pendulum response is usually simplified by considering the input from uni-axial translational motion only. Complete ground motion however, includes not only translational components but also rotations (tilt and torsion). We consider complete equations of motion for three following types of pendulum: (i) conventional mass-on-rod, (ii) mass- on-spring type, and (iii) inverted (astatic), then their response sensitivities to each component of complex ground motion are examined. Inverted pendulums are used in seismology for more than 100 years, for example, classical Wiechert's horizontal seismograph built around 1905 and still used at some seismological observatories, and recent Guralp's horizontal seismometers CMG-40T and CMG-3T. Inverted pendulums also have significant importance for engineering applications where they are often used to simulate the dynamic response of various structural systems. The results of this study show that a horizontal pendulum similar to a modern accelerometer used in strong motion measurements is practically sensitive to translational motion and tilt only, while inverted pendulum is sensitive not only to translational components, but also to angular accelerations and tilt. For better understanding of the inverted pendulum's dynamic behavior under complex ground excitation, relative contribution of each component of motion on response variants is carefully isolated. The responses of pendulums are calculated in time-domain using close-form solution Duhamel's integral with complex input forcing functions. As compared to a common horizontal pendulum, response of an inverted pendulum is sensitive to acceleration of gravity and vertical acceleration when it reaches the level close to 1.0 g. Gravity effect introduces nonlinearity into the differential equation of motion, and results in shift of the frequency response to lower frequencies. The equations of inverted pendulum represent elastic response of pendulums (as material behavior), with nonlinearity created by time and amplitude dependence of equation coefficients. Sensitivity of inverted pendulum to angular acceleration of tilt is proportional to the length of a pendulum, and should be taken into consideration since it can produce significant effect especially for long pendulums, idealizing for instance, bridge piers, bents, elevated water tanks, telecommunication towers, etc.
S43B-1891
Impacts of Rotational Ground Motions on Structural Systems
Current engineering practice establishes seismic demands based on horizontal component of ground motion; rotational (tilting) and vertical components are almost always neglected. However, ground rotations can be significant in the near-field zone. For example, tilt reached more than 3 degrees at the Pacoima Dam upper left abutment during the 1994 Northridge Earthquake. Coupling of rotational and translational components of ground motion may impose amplified seismic demands on structures. To consider these enhanced demands in design and performance assessment, governing equation of motion for multi-component (horizontal, vertical, and rotational) excitation is presented. The expanded equation includes an additional overturning moment at the base of the inverted oscillator created by gravitational acceleration on the oscillator's mass when it is displaced from its original position. This effect called P-Delta in engineering is originated by ground tilting and vertical acceleration in addition to inertia forces due to angular and horizontal accelerations. Using an idealized model of a bridge bent, inelastic structural response parameters are computed considering single and multi-component excitation; and it is demonstrated that higher ductility demand may ensue when vertical and rotational components are coupled with horizontal component. Ignoring this coupling may lead to significant underestimation of seismic demands produced by earthquake ground shaking.
S43B-1892
Observation of Transient Measurement with a Silica long base Tiltmeter
Tilt measurements allow us to observe the crustal flexure which is produced by a large range of loading processes, such as hydrological or atmospheric loadings, or tectonic loading or the weight of the moving water column during tsunamis. Proposing a detailed process at the origin of transient events requires much more minute observations than are available today. Most of the predictions for the amplitude of these transient signals point out that very weak signals must be indeed detected. It is ever more difficult for the smallest events which could be numerous at depth. An instrumental resolution much better than 0.1 µrad is required at least for the strain. This is clearly out of reach for both GPS and inSAR approaches. Our objective is to obtain tiltmetric data with a resolution better or equal than 0.001 microrad, and long term precision ranging from 0.1 microrad to 0.01 microrad by year. We have developed a silica long base tiltmeter to study hydrological load or pressure effect and to study the seismic cycle. We show the results of two 100 m silica water tube tiltmeters which were installed in a mine in the French Vosges massif in the framework of a hydrology research project and too tectonic research. This instruments show a remarkably good stability (0.0065 microrad by month) and a low noise level (of the order of 0.00001 microrad). Toroidal and speroidal free modes of the Earth were observed after the two last major events on Sumatra. In the same project we installed 4 others tiltmeters stations. In the Morbihan to study the confined aquifer of Ploemeur or to study the hydrological Karst effect in the Larzac area (Herault) or Calern (Alpes Maritimes). The Last tiltmetric station is in the North Chile in the seismic gap zone. On this station we are able to observe the tsunamis wave produced by the last Tocopilla earthquake (M7.8), many cosismic signal and we think, we are able to observe the tectonic load.