104 research outputs found
3D Gorgon Shear Wave Velocity Model
No full textThis is the final 3D shear wave velocity model derived from the 3D interpolation of several 2D shear wave velocity models produced as a part Chen & Saygin (2022)
Deep Crustal Seismic Reflection Profiling: Australia 1978-2011
No full textDeep Crustal Seismic Reflection Profiling: Australia 1978–2015 presents the full suite of reflection profiles penetrating the whole crust carried in Australia by Geoscience Australia and various partners.
The set of reflection data comprises over 16,000 km of coverage across the whole continent, and provides an insight into the variations in crustal architecture in the varied geological domains. Each reflection profile is presented at approximately true scale with up to 220 km of profile per page and overlap between pages. Each reflection section is accompanied by a geological strip map showing the configuration of the line superimposed on 1:1M geology. The compilation includes a suite of large-scale reflection transects groups of 1,000 km or more that link across major geological provinces, and an extensive bibliography of reports and relevant publications
January 2022 Arthur River M 4.8 Earthquake - Recorded by Perth DAS cable
No full textThese are the data files (SAC) of the Arthur River 24 January 2022, 21:24:47 (UTC) earthquake. Each SAC file contains coordinate information of the sensing point with duration minimum expected P wave arrival time - 20s (tp is uniform for all channels). The data was downsampled to 100 Hz. \nLineage: Data was produced via connecting CSIRO owned iDAS v2 DAS interrogator to the unused telecom cables located in the AARC building - Level 1, Kensington, WA
AuSeis: A seismic model for the Australian crust obtained from the inversion of teleseismic P-wave coda autocorrelation
No full textAuSeis makes use of the teleseismic waveforms recorded on the vertical components of 1200 permanent and temporary seismic stations across Australia. It includes the first-order estimates of the crustal Vp, Vs, Vp/Vs, density and Moho depth. It has been obtained by inverting the autocorrelations of teleseismic P-wave coda through a Markov Chain Monte Carlo approach. For further details, please refer to the paper referenced above (https://doi.org/10.1029/2018JB017055).\n\nThe results of the inversion are provided as ASCII files (.dat and .xyz files). The zip file containing the results includes the following:\n\n“Best_2000_Models” includes the best 2000 accepted 1-D models for each seismic station. The models give the thickness, depth, Vp, Vs and Vp/Vs of each crustal layer. For each station, the best 2000 accepted horizontal slowness value (s/km) are also given in the last column of each file. This parameter can be particularly useful to create synthetic seismograms associated with each of the models provided in the files.\n“1-D Models” includes the average of 1-D crustal properties (Vp, Vs, density, depth, Vp/Vs, and Moho) and their associated one sigma uncertainty for each seismic station. The mean and one sigma uncertainty for each of the crustal properties are calculated from the best 2000 accepted models.\n“AuSeis_Crustal_Model” includes an ASCII grid file (“AuSeis_Crustal_Model.xyz”). The file includes grids of seismic crustal properties (Vp, Vs, Vp/Vs, density) and their related one sigma uncertainties. The grid interval in the x and y directions is 0.5 degree, and the depth interval is 5 km. More information about the columns are provided in the file at the top of the file.\n“AuSeis_Moho” includes a file (“AuSeis_Moho_From_1D_Models.dat”) containing Moho depths and their one sigma uncertainty.\n“Stations_Used.dat” is a list of all stations used in this study. The file is organised with the following format: name of the seismic station, name of the seismic network, longitude (degree), and latitude (degree).\nLineage: It has been obtained by inverting the autocorrelations of the teleseismic P-wave coda through a Markov Chain Monte Carlo approach. For further details, please refer to the paper referenced above (https://doi.org/10.1029/2018JB017055)
Rayleigh Wave Dispersion Spectrum Inversion Across Scales
No full textTraditional approaches of using dispersion curves for S-wave velocity reconstruction have limitations, principally, the 1D-layered model assumption and the automatic/manual picking of dispersion curves. At the same time, conventional full-waveform inversion (FWI) can easily converge to a non-global minimum when applied directly to complicated surface waves. Alternatively, the recently introduced wave equation dispersion spectrum inversion method can avoid these limitations, by applying the adjoint state method on the dispersion spectra of the observed and predicted data and utilizing the local similarity objective function to depress cycle skipping. We apply the wave equation dispersion spectrum inversion to three real datasets of different scales: tens of meters scale active-source data for estimating shallow targets, tens of kilometers scale ambient noise data for reservoir characterization and a continental-scale seismic array data for imaging the crust and uppermost mantle. We use these three open datasets from exploration to crustal scale seismology to demonstrate the effectiveness of the inversion method. The dispersion spectrum inversion method adapts well to the different-scale data without any special tuning. The main benefits of the proposed method over traditional methods are that (1) it can handle lateral variations; (2) it avoids direct picking dispersion curves; (3) it utilizes both the fundamental and higher modes of Rayleigh waves, and (4) the inversion can be solved using gradient-based local optimizations. Compared to the conventional 1D inversion, the dispersion spectrum inversion requires more computational cost since it requires solving the 2D/3D elastic wave equation in each iteration. A good match between the observed and predicted dispersion spectra also leads to a reasonably good match between the observed and predicted waveforms, though the inversion does not aim to match the waveforms.We appreciate Caroline Johnson’s help in improving the manuscript. E. Saygin and L. He wish to acknowledge financial assistance provided through Australian National Low Emissions Coal Research and Development (ANLEC R&D). E. Saygin was supported by CSIRO’s Deep Earth Imaging Future Science Platform. L. He was supported by China Scholarship Council. ANLEC R&D is supported by Australian Coal Association Low Emissions Technology Limited and the Australian Government through the Clean Energy Initiative. We thank the InterPACIFIC project for providing the field data (data available here: http://interpacific.geopsy.org/). We thank David Lumley for his contribution to the retrieval of the continuous part of the SW HUB dataset (data available here: https://wapims.dmp.wa.gov.au/WAPIMS/Search/SwHubCarbonStorage). We thank Weisen Shen for providing the reference S-wave velocity model (data available here: http://ciei.colorado.edu/Models/). We extend our thanks to Roman Pevzner for providing the reference S-wave velocity of Harvey-1 well. Z. Zhang and T. Alkhalifah thank KAUST for its support and specifically the seismic wave analysis group members for their valuable insights. For computer time, this research used the resources of the Supercomputing Laboratory at King Abdullah University of Science and Technology (KAUST) in Thuwal, Saudi Arabia
Dataset and 3D Vs Model for "Crustal velocity images of north-western Türkiye along the North Anatolian Fault Zone from transdimensional Bayesian ambient seismic noise tomography"
No full text<b>External Organisations</b><br/>Istanbul Technical University; Zhejiang University; University of Leeds<b>Associated Persons</b><br/>Buse Turunctur (Creator); Tuna Eken (Creator); Yunfeng Chen (Creator); Tuncay Taymaz (Creator); Gregory A. Houseman (Creator)Final 3D Vs model and dispersion data for the paper entitled "Crustal velocity images of north-western Türkiye along the North Anatolian Fault Zone from transdimensional Bayesian ambient seismic noise tomography". In the vel_files folder, there are 10 files for each depth for 1-15 km. The format of each velocity file is as follows: Column Value 1 Lattitude (°) 2 Longitude (°) 3 Vs (km/s) The format of the dispersion data is as follows (See Computer Programs in Seismology Tutorials - do_mft for more information on the format): Column Value 1 Type of file, MFT96 2 Wave type: R for Rayleigh 3 Dispersion type: U for group velocity 4 Mode: 0 represents the fundamental mode 5 Filter period, T, in seconds 6 Dispersion value, either group or phase 7 Error in dispersion. This is just a place holder since there is no way to estimate an error from a single trace. The group velocity error is determined from the ratio of the filter period to travel time 8 Distance in km 9 Azimuth from the source to the receiver 10 Spectral amplitude. 11 Epicenter latitude 12 Epicenter longitude 13 Station latitude 14 Station longitude 15 control flag 16 control flag 17 Instantaneous period if this is preferred. This differs from the ilter period because the signal spectram is not flat. 18 Comment: keyword 19 Station 20 Component 21 Year 22 Day of year 23 Hour 24 Minute - these identify the event origin tim
Seismic Inversion by Hybrid Machine Learning
No full textWe present a hybrid machine learning (HML) inversion method, which uses the latent space (LS) features of a convolutional autoencoder (CAE) to estimate the subsurface velocity model. The LS features are the effective low-dimensional representation of the high-dimensional seismic data. However, no equations exist to describe the relationship between the perturbation of an LS feature and the velocity perturbation. To address this problem, we use automatic differentiation (AD) to connect the two terms. Following this step, we use the wave-equation inversion to invert the LS features for the subsurface velocity model. The HML misfit function measures the LS feature differences between the observed and predicted seismic data in a low-dimensional space, which is less affected by the cycle-skipping problem compared to the waveform mismatch in a high-dimensional space. A low dimensional LS feature mainly contains the kinematic information of seismic data, while a large dimensional LS feature can also preserve the dynamic information of seismic data. Therefore, the HML inversion can recover the subsurface velocity model in a multiscale approach by inverting the LS features with different dimensions. Based on the different ways of utilizing AD to compute the velocity gradient, we propose full- and semi-automatic approaches to solve this problem. These two approaches are mathematically equivalent; the former is easier to implement, while the latter is computationally more efficient. Numerical tests show that the HML inversion method can effectively recover both the low- and high-wavenumber velocity information by inverting the LS features with different dimensions.This research was fully funded by the Deep Earth Imaging Future Science Platform, CSIRO. The authors thank Dr. Mehdi Tork Qashqai and Dr. Caroline Johnson for reviewing an earlier version of the manuscript. They would like to thank Dr. Ben Harwood and Dr. Muming Zhao from Data 61, CSIRO for their guidance and insights on convolutional autoencoder. They would like to thank the Center for Subsurface Imaging and Fluid Modeling (CSIM), KAUST for the release of the Aqaba data. They also appreciate the time and efforts of the editor - Prof. Yehuda Ben-Zion, associate editor - Prof. Andrew Curtis, reviewer - Prof. Tariq Alkhalifah and one anonymous reviewer in reviewing this manuscript
Ambient seismic noise tomography of Australian continent
No full textImaging using information derived from the cross-correlation of the ambient seismic wavefield at different stations has recently become an important tool in seismology. We here present a continent wide study of the Australian crust based on the exploitation of continuous data from extensive portable broad-band deployments across Australia and the permanent stations. Permanent stations play a valuable role in linking the information from different portable deployments. Over 2000 Rayleigh wave components of the Green's functions are extracted from the inter-station cross-correlations and provide a reasonably uniform sampling of the continent. Rayleigh wave group velocities are extracted for the period range from 5 s to 12.5 s. The group dispersion from the various paths are inverted to produce group wavespeed maps based on a 2° × 2° grid using a nonlinear-iterative 2-D tomographic scheme with updating of propagation paths using the fast marching method. The group wavespeed maps display prominent features with lowered wavespeeds. For the shortest periods these features are mostly associated with the regions of thick sedimentary sequences, such as the Amadeus and Officer basins in central Australia. At the longer periods reduced wavespeeds are most likely due to elevated temperatures and link well to estimates of crustal heat flow. The major cratonic blocks show faster group wavespeeds, and the Archaean cratons of Western Australia are particularly fast with some indication of internal structure linked to terrane boundaries. The transition from the Precambrian core of the continent, in the centre and west, to the Phanerozoic fold belts in the east is not marked by any single well-defined anomaly in the crust, even though distinctive contrasts have been mapped in the mantle lithosphere from surface wave tomography
- …
