Senior Research Associate - MARS / IceDice
About the role
The IceDice project aims to provide reliable probabilistic forecasts of the West Antarctic Ice Sheet’s contribution to future sea level rise – information of enormous societal and economic value for coastal planning and climate adaptation worldwide.
Your research will develop and apply novel Bayesian machine learning methods – in particular physics-informed Gaussian processes and/or neural operators– to build accurate probability density functions (PDFs) of future Antarctic sea level contributions.
Working as part of a collaborative team spanning British Antarctic Survey and the University of Cambridge, you will be based at Lancaster and work with Dr Henry Moss to:
- Develop machine learning emulators for the WAVI ice-sheet model to serve as efficient surrogates for large-scale Bayesian inference.
- Develop utility-function-based experimental design methods to identify the computer simulations and observational surveys that maximise information about future sea level for a given computational cost.
- Work closely with ice-sheet modellers at the British Antarctic Survey to apply probabilistic methods to realistic West Antarctic domains.
- Publish high-quality research in leading peer-reviewed journals and present at national and international conferences.
What we’re looking for
- PhD in statistics, machine learning, physics or a closely related discipline.
- Research experience in Bayesian methods, probabilistic modelling, or scientific machine learning.
- Experience with Gaussian processes, MCMC methods, or uncertainty quantification for expensive computational simulators.
- An interest in applying mathematical methods to real-world environmental challenges, and willingness to collaborate across disciplinary boundaries.
- Experience with Python, or equivalent scientific computing languages.
This is a full-time, fixed-term position for 22 months or until 31/08/2028 (whichever comes first). Flexible working arrangements will be considered, but you will be expected to be present on the Lancaster campus a minimum of two days per week.
Find Your Best Opportunity
Tell them AcademicJobs.com sent you!
