Abstract: We measured shear wave anisotropy under the Geometry of Cocos (GECO) experiment, deployed at the eastern end of the Trans-Mexican Volcanic Belt. A tear in the south Cocos slab has been proposed [Dougherty and Clayton, 2014]. Two splitting parameters are used to quantify anisotropy. These are the delay time (delta t) and the fast polarization direction (phi). The shear wave splitting method of Silver and Chan [1991] was used. A time window containing the SKS or PKS phase is selected. Then, the north-south and east-west components are rotated by 1 degree, ranging from -90º to 90º. For each rotation, the solution is searched in 0.05s increments, and the crosscorrelation and autocorrelation between the two traces is obtained. Next the eigenvalues are calculated for every possible combination of (delta t) and (phi). Finally, the minimum eigenvalue is chosen as the best solution. We check our results in two ways: by correcting the original records using the measured (delta t) and (phi) in order to obtain a weaker SKS signal in the transverse component, and by comparing the waveforms and the arrival time difference between the fast and the slow components. The orientation of (phi), measured for western stations, is explained by subslab, entrained mantle flow under flat Cocos slab. While going from west to east. the slab geometry changes from flat to steeper subduction, and may be accompanied by a possible tear in the slab. For the eastern stations, (phi) is rotated 25º clockwise relative to stations in the west. This phenomenon may be accounted for by mantle flow between the possible tear and by rollback of the slab. The orientation of the fast axes where the slab subducts steeply is consistent with both subslab entrained flow, and with corner flow in the mantle wedge between the Cocos and North American plates.
Slidecast:
https://vimeo.com/276938525
Abstract: The distinct sensitivities of seismic velocity and attenuation (1/Q) to water content, melt, and major element composition yield important constraints on mid-ocean ridge processes and the associated mantle flow pattern, melt distribution, and the interaction of spreading centers with hotspots. Beneath the Juan de Fuca ridge, both body-wave and surface-wave studies observed strong attenuation and slow velocity anomalies. However, it is still under debate which mechanism is favored: Either the direct effects of in situ melt extending to depths of 150 km or more, or the pre-melting effects of a hydrous mantle upwelling with about 200 ppm of water in which melting commences at about 100 km. Neither the depth-integrated S-wave attenuation nor the averaged 1-D shear attenuation for surface waves can provide further constraints; to resolve the debate a 3-D attenuation structure is required. Surface-wave tomography of 3-D attenuation structure is known as a challenging task due to difficulties in separating elastic focusing effects on seismic amplitudes from intrinsic attenuation and the relatively small effects of attenuation on amplitude over short distances. Taking this into account, we introduced the finite-frequency kernels for attenuation and imaged 3D shear attenuation and velocity structures in the vicinity of the Juan de Fuca plate. Owing to the extensive arrays of ocean bottom seismometers (OBS) with high quality, long deployment periods, and novel noise removal techniques, we obtained 3-D attenuation and shear velocity model with the best resolution to date of any spreading center based on fundamental-mode Rayleigh waves. With 3-D models of shear attenuation and shear velocity, we evaluate the two competing paradigms and shall present a summary of the inversion results and favored interpretation.
Slidecast:
https://vimeo.com/276930248
Abstract: Faults are surrounded by a hierarchical structure of damaged rocks. Such damaged fault zones can extend several hundred meters across major faults such as San Andreas, Nojima and North Anatolian faults. They also show along-strike variation of damage that may be correlated with historical earthquake ruptures. Due to the change of energy required for rupture propagation, the variation of fault zone structure can either prohibit or assist rupture propagation and significantly affect the final sizes of earthquakes. Recent dense array data recorded at the San Jacinto fault zone suggests the existence of three prominent fault zones across the Anza seismic gap and the south section of the Clark branch, whereas fault zones were not detected near the ends of the Anza seismic gap. We investigate the effects of along-strike variation of fault zone damage using dynamic rupture simulations, which calculate the time-varying rupture process by considering the interactions between fault stresses, friction, and material heterogeneities. We first focus on along-strike rupture propagation in a 2D configuration. We will show that, for small nucleation sizes and short rupture propagation distances, ruptures nucleated in fault zones tend to terminate when they enter intact rocks outside fault zones. The stopping effect is less pronounced when fault zones become wider, sharper and more damaged, indicating a temporal correlation between the fault zone structure and break-through ruptures. Moreover, break-through ruptures are also expected when a sufficiently-large high-stress asperity exists at the end of the fault zone. We will then present 3D scenarios of San Jacinto earthquake ruptures and investigate whether ruptures can break through the Anza seismic gap given the current stress state and the distribution of fault zones. Our results suggest that a priori knowledge of the fault zone structure is of great importance for predicting sizes of future large earthquakes on major faults.
Slidecast:
https://vimeo.com/276934713
Abstract: Seismic approaches based on the phase/traveltimes are widely and successfully used to image the subsurface geological structure. On the other hand, the amplitude of seismic waveform that is hoping for providing more physical information about geology has not been explored in a consistent way. There exists a gap between current seismic capability and the full physics of seismic waveform. For example, as Earth media always attenuate seismic waves during propagation seismic data includes attenuation effects that contributes seismic amplitude seriously. To fill the gap, we have to deal with seismic attenuation (intrinsic and scattering) physically and practically in next-generation seismic techniques. In this talk, I will review recent progress on the development of theory of viscoacoustic and viscoelastic wave equations to model wave attenuation in the Earth. The theory is based on the frequency-independent Q model, but it can be generalized to frequency-dependent Q. Then I’ll show how this wave equation facilitates the compensation of attenuation effects in seismic imaging in the context of time-reversal seismic imaging, reverse-time migration (RTM), and full waveform inversion using synthetic data and field data. Finally, I will discuss the prospective research in seismic attenuation modeling, imaging, and full waveform inversion.
Slidecast:
https://vimeo.com/276930129
Abstract: We have investigated seismicity associated with hydraulic fracturing (HF) in Ohio since 2013, which provides an ideal setting for studying the relations between high pressure injection and earthquakes due to isolation from other injection activities. Our analysis using an array of local stations in Harrison County revealed 2 distinct groups: 1) deeper earthquakes in Precambrian basement, with larger magnitudes (M>2), b-values < 1, and many post shut-in earthquakes, versus 2) shallower earthquakes in Paleozoic rocks ~400 m below HF, with smaller magnitudes (M<1), b-values > 1.5, and few post shut-in earthquakes. Based on geologic history, laboratory experiments, and fault modeling, we interpreted the deep seismicity as slip on mature faults in older crystalline rocks and the shallow seismicity as slip on immature faults in younger sedimentary rocks. Wells inducing deeper seismicity produced more water than wells with shallow seismicity, indicating more extensive hydrologic connections outside the target formation, consistent with pore pressure diffusion influencing seismicity. However, the 2-3 hours between onset of HF and seismicity is too short for typical fluid pressure diffusion rates across distances of ~1 km and argues for poroelastic stress transfer also having a primary influence on seismicity. We now extend our analysis to other cases of HF induced seismicity in the Appalachian Basin. While these cases did not have publicly available local arrays, a broader set of stations operated by Miami University, ODNR, USGS, and IRIS is sufficient to identify the primary patterns of these seismic sequences. We employ multistation template matching to improve detection, waveform correlation to improve phase arrivals, and double difference relocation to improve the hypocentral characterization. We also examine frequency-magnitude distributions and well production patterns to compare with the geologic and operational interpretations made in the Harrison County study.
Slidecast:
https://vimeo.com/276939431
Abstract: Rapid detection and classification of damage after earthquakes is important for loss estimation, rapid response, and research. We use optical satellite imagery after the recent M 7.1 Central Mexico earthquake on September 19, 2017 to develop a damage catalog for Mexico City. The satellite imagery before and after the earthquake are freely available through the Digital Globe disaster program with a spatial resolution of less than 1-meter. The pre-event image was taken on June 15, 2017 through GeoEye-1 sensor and the post-event image was taken on September 20, 2017 through WorldView-2 sensor. Machine learning algorithms (MLAs) including feed-forward and radial basis artificial neural networks (ANNs) are used to detect and classify building damage after the earthquake. MLAs work with nonlinear datasets, learn from limited training data, and have been successfully used in other classification problems. In addition to the spectral information of the imagery, textural and structural features such as dissimilarity and Laplacian of Gaussian (LoG) filter are used as the inputs to the MLAs and result in an improvement of the overall accuracy of the classification. Terrestrial images taken by individuals from the damaged buildings after the earthquake, which are available online and through reconnaissance reports, are used as ground truths to develop both training and testing data for the region. There are more than 1000 buildings that are partially damaged while around 20 are reported as collapsed. As the spatial and spectral resolution of the imagery are not high enough to detect the partial damages to the buildings, this study only focuses on totally collapsed structures. The classification results are validated using 2-fold cross-validation with a confusion matrix and to evaluate the overall accuracy of the algorithms. The results of this work provide preliminary evidenced that collapse catalog using MLAs and high resolution (1-m) optical imagery can be developed to inform loss estimation, rapid response and research efforts after major earthquakes.
Slidecast:
https://vimeo.com/276977582
