PhD Studentship - Data Driven Runout Modelling to Improve Coseismic Landslide Hazard Forecasts

About the Project

Overview of Project

This project develops next‑generation models to predict where earthquake‑triggered landslides will occur and how far they will travel; critical information for protecting lives, infrastructure, and emergency responders. By combining global landslide data, innovative machine‑learning methods, and new ways of representing runout, the research will produce faster and more reliable nowcasts for use immediately after major earthquakes. Working with the Sichuan Earthquake Authority ensures that the tools address real operational challenges. The outcomes will strengthen disaster‑response capabilities worldwide and support long‑term resilience planning in regions facing severe seismic hazard.

Research questions

1. How can runout processes be represented in a computationally efficient, data-driven framework suitable for rapid post-earthquake nowcasting?
2. To what extent does integrating runout improve predictive skill relative to current initiation-only models used by agencies such as the Sichuan Earthquake Authority and USGS?
3. Which combinations of global and locally sourced predictors (e.g., topography, lithology, shaking metrics) most improve performance, and how transferable are these improvements across diverse earthquake events and landscapes?
4. How does model skill vary between pre- and post-event settings, particularly under uncertain shaking scenarios in pre-event assessments?
5. What level of model complexity provides the optimal trade-off between physical realism, generalisability, and computational speed for operational disaster-response contexts?

Methodology

The research will begin by compiling and expanding a global library of co-seismic landslide inventories, including new datasets generated using the ALDI automated detection algorithm (Milledge et al., 2022) and subsequently validated through manual refinement. Predictor variables will be assembled from global datasets (e.g., USGS shaking metrics, global DEMs, lithology maps) and from high-resolution local sources held by regional agencies such as the Sichuan Earthquake Authority. Model development will combine empirical machine-learning approaches with physics-informed constraints derived from non-local hillslope transport theory (Furbish et al., 2020). Existing lightweight runout formulations (Milledge et al., 2019) will be extended and benchmarked against detailed physical models such as RAMMS (Zimmerman et al., 2020) and MassWastingRunout (Keck et al., 2024) in controlled test cases. A suite of initiation-runout coupled models will then be trained and cross-validated using diverse landscapes to minimise overfitting. Performance will be evaluated using hold-out testing on both within-event and cross-event datasets, with comparison against current state-of-the-art systems used operationally. Particular emphasis will be placed on runtime, interpretability, and suitability for integration into real-time earthquake information platforms.

Anticipated outcomes & benefits for the sponsoring organisation and other stakeholders

The project will deliver two operationally relevant models: a fast post-event nowcasting model suited to emergency response and a higher-fidelity pre-event model for hazard planning. Both will improve predictive skill by incorporating runout explicitly and by leveraging diverse global and regional datasets. For sponsoring organisations, including the Sichuan Earthquake Authority, the research will provide actionable, rapidly deployable hazard information, enabling better prioritisation of rescue routes, infrastructure assessments, and resource allocation immediately after major earthquakes. For other international stakeholders such as the USGS, humanitarian agencies, and national disaster-management authorities, the project will offer transferable tools and quantitative insights into how coupling initiation and runout enhances hazard forecasts. More broadly, the work will advance scientific understanding of non-local landslide processes and support the development of globally consistent, operational landslide nowcasting frameworks.

Start Date

21st September 2026

Duration of Award

3.5 years

Sponsor

NERC

Eligibility Criteria

A 2:1 Honours degree, or international equivalent, in a subject relevant to the proposed PhD project (inc. geography, earth sciences, engineering, computing and mathematics)

The project does not require the student to arrive with specialist technical expertise. Most of the skills needed (e.g. geospatial analysis, data handling, programming, and landslide‑hazard concepts) can be developed during the PhD with support from the supervisory team and training available through i‑Risk and Newcastle University. A willingness to engage with quantitative and computational approaches is the key requirement. Some familiarity with geospatial data or basic programming would be helpful, but these are not essential. The project is well suited to candidates from earth science, geography, engineering, or related backgrounds who are motivated to learn new technical skills as part of their doctoral training. Overall, the key competencies, working with spatial datasets, learning modelling techniques, and developing an understanding of landslide processes, are highly trainable, and the supervisory team is experienced in supporting students in acquiring them throughout the PhD.

Home and international applicants (inc. EU) are welcome to apply and if successful will receive a full studentship. Applicants whose first language is not English require an IELTS score of 6.5 overall with a minimum of 5.5 in all sub-skills.

International applicants may require an ATAS (Academic Technology Approval Scheme) clearance certificate prior to obtaining their visa and to study on this programme.

How to Apply (Web Text)

You must apply through the University’s Apply to Newcastle Portal

Once registered select ‘Create a Postgraduate Application’.

Use ‘Course Search’ to identify your programme of study:

• search for the ‘Course Title’ using the programme code: 8040F
• select ‘PhD Civil Engineering – Civil Engineering (Water Resources) (full time)’  as the programme of study

You will then need to provide the following information in the ‘Further Details’ section:

• a ‘Personal Statement’ (this is a mandatory field) - upload a document or write a statement directly into the application form
• the studentship code IRISK03 in the ‘Studentship/Partnership Reference’ field
• when prompted for how you are providing your research proposal - select ‘Write Proposal’. You should then type in the title of the research project from this advert. You do not need to upload a research proposal.