Seismic Imaging of Earth Across Scales
This project is about developing next-generation seismic imaging methods using the elastic energy originating from earthquakes, random vibrations and human-made sources. The ultimate goal is to improve our understanding of the Australian and surrounding regions' subsurface structure, which will help to pinpoint the location of the natural resources
Area of science
Geophysics, Geosciences
Systems used
Magnus, Zeus and Managed Storage
Applications used
In-house developed codes, Python, Fortran and C compilersThe Challenge
The majority of the Australian continent is covered by thick sedimentary basins, which limits our understanding of the subsurface beneath these structures. Traditionally, a large portion of Australian mineral commodities have been mined at locations, where there was an outcrop or the mineral deposit was close to the surface.
Because it is still not known what lies beneath this thick sedimentary cover, the mineralisation potential beneath these basins is poorly understood. To increase our understanding, one potential solution is to conduct systematic drilling, which is often prohibitively expensive and can be done up to depths of 15 km. Another solution is to image the subsurface using non-invasive advanced geophysical imaging techniques
The Solution
We as Seismic Imaging Theme of Deep Earth Imaging Future Science Platform (SI-DEI FSP) of CSIRO develop next generation of subsurface seismic imaging algorithms coupled with Bayesian modelling & Machine Learning. We then use these new methods with active and passive seismic wavefield to image and understand Australian and surrounding regions’ subsurface structure across scales
The Outcome
The following new methods were developed in this process by the members of SI-DEI FSP:
-Subsurface structure of the Australian Continent from a new method using Chaotic Part of Earthquake Waves (Tork Qashqai et al., 2019).
-New Bayesian Seismic Imaging Method (Guo et al., 2020; Visser et al., 2019).
-New Passive Seismic Reconstruction and Imaging Method (Chen & Saygin, 2020)
-New Machine Learning Based Seismic Inversion Method (Chen at el., 2020)
-New Method for revealing Seismic Waves Buried deep in the signal (Saygin & Kennett, 2019).
List of Publications
Chen, Y., Saygin, E. (2020). Empirical Green’s Function Retrieval using Cross-correlation of Ambient Noise Correlations (C2), J. Geophys. Res.-Solid Earth, doi:10.1029/2019JB018261.
Chen, Y., E. Saygin and G. T. Schuster, 2020, seismic inversion by multi-dimensional Newtonian machine learning: 90rd Annual InternationalMeeting, SEG, Expanded Abstracts (submitted).
Guo, P., Visser, G., and Saygin, E. (2020). Bayesian trans-dimensional reflection full waveform inversion: synthetic and field data application, Geophys. J. Int., accepted.
Saygin, E., Kennett, B.L.N, (2019). Retrieval of Interstation Local Body Waves from Teleseismic Coda Correlation, J. Geophys. Res.-Solid Earth, doi:10.1029/2018JB016837.
Tork Qashqai, M., Saygin, E., Kennett, B.L.N. (2019). Crustal Imaging with Bayesian Inversion of the Teleseismic P-wave Coda Autocorrelation, J. Geophys. Res.-Solid Earth, doi:10.1029/2018JB017055.
Visser, G., Guo, P., Saygin, E. (2019). Bayesian Transdimensional Seismic Full Waveform Inversion with a Dipping Layer Parameterization, Geophysics, doi:10.1190/geo2018-0785.1.
