Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Full-waveform inversion for imaging the Pacific mid-mantle seismic structure
This master thesis project aims at deriving a tomographic model of the deep interior seismic structure of the Pacific area. The methods are based on spectral-element wave propagation, full-waveform inversion, data processing and numerical optimisation.
Keywords: Computation and Global seismology; Full-waveform inversion; Inverse theory; Adjoint methods; Data processing; Earth's deep interior; Finite-frequency theory; Body and Surface waves.
One of the major interests in the mid-mantle under the Pacific is the occurrence of seismic heterogeneities, suggesting lateral variations of multi-scale nature that could explain thermal and compositional variations in the earth’s material flow. So far, seismic tomography models may lack the necessary resolution to reliably connect seismic features to these properties. Much effort from researchers has added a lot of knowledge and with presenting a new approach based on FWI, we aim at contributing to increasing valuable information on velocity variations and heterogeneous structures in Pacific’s mantle.This project’s methodology is based on spectral-element wave propagation and adjoint methods in order to obtain updated models within the framework of CSEM. We compute adjoint sensitivity kernels to update elastic model parameters, namely Vp and Vs. These kernels capture finite frequency effects and wave complexities, which gives an advantage compared to existing images of the Pacific deep interior. Advances in spectral-element modelling give us the possibility to also include effects of oceanic load during modelling of waveforms and potentially improve resolution, in case free-surface reflected or surface waves are considered.Leveraging the increase in seismic waveforms obtained from flexible arrays in the peripheral continents, the student will be able to conduct investigations of diffracted and converted body waves in low-to-intermediate frequencies and coda of the P and S waves. These have been useful datasets and testing their traveltime and amplitude sensitivity within FWI schemes would be useful for progressing the imaging of mid-mantle. Initial model updates using gradient descent methods will be able to illuminate low-to-intermediate wavelength structures. Additional steps may include imaging of heterogeneities clustering in various regions, relating to upwelling hot material or downwelling subtuction slab material.The prospective student working on the project will have the opportunity to gain experience with spectral-element wave propagation, optimal waveform data processing and waveform-to-structure in-version methods using FWI. Additional to mastering the technical aspects of the project, the student will have the chance to explore the role of seismological data in unraveling the not fully understood processes that control surface observed phenomena, happening inside our dynamic planet.
One of the major interests in the mid-mantle under the Pacific is the occurrence of seismic heterogeneities, suggesting lateral variations of multi-scale nature that could explain thermal and compositional variations in the earth’s material flow. So far, seismic tomography models may lack the necessary resolution to reliably connect seismic features to these properties. Much effort from researchers has added a lot of knowledge and with presenting a new approach based on FWI, we aim at contributing to increasing valuable information on velocity variations and heterogeneous structures in Pacific’s mantle.This project’s methodology is based on spectral-element wave propagation and adjoint methods in order to obtain updated models within the framework of CSEM. We compute adjoint sensitivity kernels to update elastic model parameters, namely Vp and Vs. These kernels capture finite frequency effects and wave complexities, which gives an advantage compared to existing images of the Pacific deep interior. Advances in spectral-element modelling give us the possibility to also include effects of oceanic load during modelling of waveforms and potentially improve resolution, in case free-surface reflected or surface waves are considered.Leveraging the increase in seismic waveforms obtained from flexible arrays in the peripheral continents, the student will be able to conduct investigations of diffracted and converted body waves in low-to-intermediate frequencies and coda of the P and S waves. These have been useful datasets and testing their traveltime and amplitude sensitivity within FWI schemes would be useful for progressing the imaging of mid-mantle. Initial model updates using gradient descent methods will be able to illuminate low-to-intermediate wavelength structures. Additional steps may include imaging of heterogeneities clustering in various regions, relating to upwelling hot material or downwelling subtuction slab material.The prospective student working on the project will have the opportunity to gain experience with spectral-element wave propagation, optimal waveform data processing and waveform-to-structure in-version methods using FWI. Additional to mastering the technical aspects of the project, the student will have the chance to explore the role of seismological data in unraveling the not fully understood processes that control surface observed phenomena, happening inside our dynamic planet.
In this master thesis project, the goal is to work towards building a full-waveform inversion(FWI) model of the Northern Pacific area, focusing on mid-mantle structure. This important region is governed by stunning geological phenomena, such as surface emergence of mantle plumes and occurrence of historical earthquakes at the rims of its major tectonic plate. These earthquakes were devastating for human societies due to their large magnitudes and consequent effects. Improving tomographic images of the region can help us to better understand and assess potential natural hazards and geological events. Seismic tomography using full waveform inversion is a powerful tool to illuminate these processes, providing high resolution snapshots at different scales.
In this master thesis project, the goal is to work towards building a full-waveform inversion(FWI) model of the Northern Pacific area, focusing on mid-mantle structure. This important region is governed by stunning geological phenomena, such as surface emergence of mantle plumes and occurrence of historical earthquakes at the rims of its major tectonic plate. These earthquakes were devastating for human societies due to their large magnitudes and consequent effects. Improving tomographic images of the region can help us to better understand and assess potential natural hazards and geological events. Seismic tomography using full waveform inversion is a powerful tool to illuminate these processes, providing high resolution snapshots at different scales.
Dr. Maria Koroni
**maria.koroni [at] erdw.ethz.ch**
Seismology and wave physics
ETH Zürich
Dr. Maria Koroni **maria.koroni [at] erdw.ethz.ch** Seismology and wave physics ETH Zürich