We are looking to appoint a Postdoctoral Research Associate to join us in an ambitious and high-impact climate research project funded by the UK’s prestigious new Advanced Research + Invention Agency (ARIA). The PROTECT project (Probabilistic Forecasting of Climate Tipping Points) brings together cutting-edge AI, statistical, and machine learning techniques with climate modelling, aiming to transform our understanding and prediction of climate tipping points.
We welcome candidates with expertise in climate modelling, ideally including experience with General Circulation Models (GCMs), Earth System Models, or related simulation tools. Additional knowledge of statistical methods, Monte Carlo techniques, or ensemble approaches is highly valued.
The post is advertised as full-time. We are open to hybrid working arrangements (on a non-contractual basis), combining on-campus and remote work.
The Opportunity:
The successful applicant will join a cross-disciplinary team working at the interface of climate science, statistical methodology, and computational modelling. The PROTECT project is part of the ARIA-funded programme on climate tipping points and aims to improve the scientific and statistical foundations of ensemble-based forecasting in climate models, mostly in general circulation models (GCMs).
You will work with the PI of the project (Prof. Victor Elvira) and collaborate with other ARIA projects. You will contribute to the following areas:
* Review and benchmark datasets used for initialization, calibration, and validation of GCMs, identifying sources of uncertainty and quantifying their impact.
* Work closely with ARIA collaborators to select appropriate GCMs or Earth System Models (e.g., CESM2, MPI-ESM, NEMO), with a focus on models relevant to AMOC and the North Atlantic.
* Develop and implement probabilistic ensemble techniques to improve uncertainty representation in climate simulations.
* Apply and validate Monte Carlo methods (e.g., importance sampling and other rare-event estimation methods) for the detection and probabilistic characterization of climate tipping points.
* Coordinate with other ARIA teams to ensure compatibility and reproducibility of methods, particularly within a shared definition and detection framework for tipping points.
* Contribute to the design of scalable and interpretable forecasting strategies for large climate simulators, integrating adaptive sampling and Bayesian techniques.
* Disseminate results in peer-reviewed publications and contribute to open-source codebases and documentation.
* Participate in the integration and delivery of a final modular framework for uncertainty-aware probabilistic forecasting, with the potential for policy relevance.
Applicants should have (or be close to obtaining) a PhD in climate science, climate modelling, geophysical fluid dynamics, or a closely related field. Experience with uncertainty-aware simulation or familiarity with statistical tools is desirable but not strictly required. Strong candidates from physical climate modelling will be seriously considered.
Your skills and attributes for success:
* Experience with climate simulation models. Ideally, experience includes running general circulation models (GCMs), Earth system models, or similar simulation tools. However, training opportunities (e.g., via NCAS) are available for motivated candidates.
* Interest in probabilistic methods, ensemble simulations, or uncertainty quantification; experience is appreciated
* Ability to engage with interdisciplinary teams involving statisticians, computer scientists, and climate experts
* Programming experience in a scientific computing language (e.g., Python, Fortran, MATLAB)
* Excellent communication skills, both written and oral
£40,497 to £48,149. Grade UE07
#J-18808-Ljbffr