Project Overview: This PhD opportunity at the University of Leeds focuses on imaging the Earth's deep interior using massive seismic datasets and advanced machine learning techniques. The project aims to unravel the complex processes of mantle convection, core-mantle interactions, and the evolution of continent-sized structures such as Large Low Velocity Provinces (LLVPs) and Ultra-Low Velocity Zones (ULVZs). These deep Earth features are crucial for understanding planetary evolution, surface processes, and the conditions that make Earth habitable.

Research Focus: You will analyze global seismic data to map heterogeneity in the lowermost mantle, applying both supervised and unsupervised machine learning approaches. The project is structured in three flexible stages: (1) mapping small-scale heterogeneity along the core-mantle boundary using seismic scattering analysis, (2) developing machine learning methods to characterize waveform variations of core-reflected seismic phases, and (3) pursuing a subproject tailored to your research interests, leveraging insights and skills gained in the first two stages.

Academic Environment: You will join the Deep Earth research group within the School of Earth & Environment, one of the largest global teams studying the Earth's deep interior. The group offers a collaborative and supportive environment, with daily interaction among supervisors, postdoctoral researchers, and fellow PhD students. International collaboration opportunities are available.

Funding: This position is part of the YES-DTN program, which provides 25–26 fully funded studentships annually. Funding covers university fees, a personal stipend, and research/training costs. International applicants are welcome, but must cover visa and health surcharge costs, and awards for non-UK applicants are limited by UKRI rules. See application information for details.

Eligibility: Suitable candidates will have a background in physics, geophysics, quantitative geology, environmental science, mathematics, or computer science. The project is ideal for those interested in applying modern data analysis and machine learning to geoscience problems. No specific GPA or language test requirements are listed, but international applicants must meet university entry and visa requirements.

Application Process: Applications are open to UK and international candidates. The deadline is January 7, 2026. Prepare your application materials and submit via the YES-DTN website. For more details, visit YES-DTN application information or the FindAPhD project page.

References: The project draws on recent advances in big data seismology and machine learning, with key references provided for further reading.