About the Project
Project Rationale:
Earth’s volatile cycle is fundamental to human existence on the planet. However, precise constraints on this system have proved challenging. The transport of volatiles into the Earth at subduction zones as well as between the lower and the upper mantle at the transition zone are key. For instance, some volatiles that enter the Earth via subduction, escape during volcanic eruptions. Some volatiles rising from the lower mantle may get trapped in dense melts just above the mantle transition zone near 410 km depth. The pathways and quantities of the volatiles are important to understand the evolution of the Earth, and represent a hot, current research topic in the Earth Sciences. This project will involve putting tight constraints on both of these systems using cutting edge techniques, like machine learning, and integration with a range of interdisciplinary constraints, for a better understanding of the inputs, pathways, and fluxes of volatiles in the Earth. There are broad implications for plate tectonics, climate change, and the habitability of the planet.
Methodology:
The student will use seismic imaging techniques including tomography, receiver functions, and SS-precursors to constrain subduction zone volatile cycling and water and melt above the transition zone. The work will begin with some key sites including for instance the mantle wedges and also transition zone structures in the Caribbean, Cascadia, Papua New Guinea, and Alaska. The student will constrain both isotropic and anisotropic structures. Then the student will use the large number of datasets and seismic velocity models that we have already achieved as templates, to perform a cutting-edge machine learning approach to the millions of waveforms that we have compiled from the global seismic data base. The student will use machine learning to cull the data and then also to invert for seismic velocity structures to constrain the pathways and quantities of volatiles in the mantle. There will also be the potential to use machine learning in a more interdisciplinary context to understand underlying relationships between different geophysical, geological and geochemical observations. The results will be fully integrated to achieve maximum impact in an interdisciplinary approach including constraints from mineral physics, geodynamic modelling, melt migration modelling, and geochemistry.
Training:
All doctoral candidates will enrol in the Graduate School of NOCS (GSNOCS), where they will receive specialist training in oral and written presentation skills, have the opportunity to participate in teaching activities, and have access to a full range of research and generic training opportunities. GSNOCS attracts students from all over the world and from all science and engineering backgrounds. There are currently around 200 full- and part-time PhD students enrolled (~60% UK and 40% EU & overseas). Specific training will include:
The student will develop skills and learn techniques from seismology as a member of one of the largest and most active geophysics groups in the UK. The student will learn to cull, process, and invert seismic data using dense seismic arrays. The student will learn and apply a wide range of cutting edge machine learning techniques to both regional and global seismic databases. The student will have excellent computational facilities and be trained in programming skills for Python, Seiscomp3, FORTRAN, Matlab, SAC, and the UNIX operating system. The student integrate the results in a broad interdisciplinary approach. A wide range of opportunities to develop the range of generic skills essential for successful completion of the PhD and their future career are available through the Graduate School NOCS including geophysical fieldwork. The project will also involve collaborative travel, especially to London, Tokyo, and the USA. This training will prepare the student for a career path in academia and industry.
References
Agius, M., N. Harmon, C. A. Rychert (2018) Tharimena, S., Kendall, M. Sediment characterisation beneath the PI-LAB experiment at the Mid-Atlantic Ridge from P-to-S conversions, Geophys. Res. Lett., doi: 10.1029/2018GL080565.
Meier et al., (2019) Reliable Real‐Time Seismic Signal/Noise Discrimination With Machine Learning, Journal of Geophysical Research, Solid Earth, 124, 788–800, doi.org/10.1029/2018JB016661
Bercovici, D. & Karato, S. I, (2003) Whole-mantle convection and the transition-zone water filter, Nature, v. 425.