PhD in Ocean and Climate Science: Modelling Abrupt Changes in Atlantic Circulation

PhD in Ocean and Climate Science: Modelling Abrupt Changes in Atlantic Circulation

Quick Decision

Promising with material uncertainties. This University of Leeds project connects three substantial tasks: build a theory for oscillating glacial simulations, test it against Dansgaard–Oeschger reconstructions from ocean sediment cores, and evaluate early-warning statistics. It is a focused match for applicants who can work across quantitative modelling, climate or ocean science, palaeoclimate data, and statistical inference. The main scientific advantage is a Leeds-led new mechanism that gives the thesis a concrete starting point. The main limitation is that its evidence comes from selected glacial simulations; it is not itself evidence that present-day AMOC is about to collapse.

The network lists a competition-funded studentship with a stipend, university fees and research/training support. Its current FAQ says international candidates may receive up to 30% of awards and that university fees are covered regardless of the home/international difference, but UK relocation expenses remain outside that support. The project page gives 13 January 2027 as its deadline; the official 2027 application instructions give 13:00 GMT on 12 January 2027. Treat the latter as binding unless YES-DTN confirms a later change. Contact the primary supervisor early, then submit both forms; the Leeds application should be in at least two days before the deadline because a student ID is needed for the second form.

Research Problem and Doctoral Work

What is unsolved and why it matters

The Atlantic Meridional Overturning Circulation (AMOC) redistributes heat through the Atlantic. Freshwater entering the North Atlantic can affect its strength, but the conditions that cause abrupt state changes, their probability, and reliable advance indicators remain uncertain. Direct instrumental observations do not cover an AMOC collapse, and projections differ. The project therefore uses the last glacial period, when abrupt Dansgaard–Oeschger (D–O) warm/cold transitions are preserved in climate records, as a test bed for mechanisms.

The advertised scientific starting point is a convection–advection oscillator proposed in a 2025 study of HadCM3 Last Glacial Maximum simulations. That work connects slower salinity/stratification changes to faster shifts in ocean convection. Its authors are explicit that the mechanism was identified in a particular simulation and may depend on the forcing, boundary conditions, model settings and model choice. The earlier simulations also show that the location and rate of meltwater forcing, as well as the background state, affect whether oscillations occur. A related 2025 team paper examines competing sea-ice effects on the pace and amplitude of those oscillations. These results give the PhD a tractable set of hypotheses, while leaving open whether the mechanism generalises to other models and records.

Work, methods and potential outputs

The official project page lists three thesis objectives:

1. Develop a theoretical framework that reproduces the oscillating simulations, moving from simple oscillator formulations to more physical box models and identifying the components responsible for AMOC shifts.
2. Compare the mechanism with D–O climate reconstructions from ocean sediment cores, including a direct test of the “global salt oscillator” hypothesis.
3. Apply statistical early-warning algorithms to numerical simulations and D–O records, then assess which indicators are robust enough to inform future AMOC monitoring.

A useful way to think about the work is as a chain of tests: can the proposed process produce the model oscillations; do the reconstructed records support the same mechanism; and do candidate indicators work when the data are incomplete, noisy, or drawn from a system whose transition is not already known? The final report should not assume that an algorithm developed on glacial simulations will predict a modern transition without further validation.

That caution is material. A 2024 critique argues that changing proxy coverage can create artificial increases in variance and false early-warning signals. A 2024 physics-based study instead proposes freshwater transport at the southern Atlantic boundary as an observable early-warning measure, while showing that conventional statistical signals are sensitive to the selected pre-transition time window in its model experiment. A 2025 review highlights diversity and biases in current Earth-system models and notes that quantitative collapse probabilities remain available only from conceptual models. Together, this evidence makes method comparison, proxy selection and uncertainty analysis central research questions—not routine validation steps.

School, Programme and Field Context

The degree-awarding institution is the University of Leeds; the project page places it in the School of Earth and Environment. University College Cork and the University of Bergen appear as co-supervision affiliations, alongside Leeds supervisors. The official project page does not describe a joint degree or a second registration, so applicants should treat Leeds as the host and ask whether any extended visits are expected.

The School's published research themes include observation and modelling of ice sheets, sea ice, ocean and atmospheric circulation, and climate sensitivity. The Physical Climate and Cryosphere cluster brings together groups in physical climate change, ice and climate, palaeoclimate and cryosphere research. This is a relevant setting for a project that must link physical mechanisms, palaeo-observation and climate modelling. In the QS World University Rankings by Subject 2026, Leeds is ranked 12th in Earth & Marine Sciences; this is a broad subject signal, not an assessment of this project or a specific ocean-climate subgroup.

The studentship funding lasts 3.5 years, while the University's standard full-time PhD period is normally three years; the funding duration should not be read as a guaranteed longer registration period. The network describes technical and scientific support from supervisors, Climate-Ice colleagues and its wider training network. The precise first-year training plan, progress/transfer requirements, model-computing allocation, data access and supervision cadence are not specified in the project advert and should be confirmed with the supervisor.

Supervisor and Lab Assessment

Dr Ruza Ivanovic is the primary supervisor and an Associate Professor in Climatology. Her official profile describes work using climate and ice-sheet models, including HadCM3 and FAMOUS, to study climate–ice–ocean interactions and abrupt change. She co-leads the Climate-Ice Research Group and leads the PMIP Deglaciations and Abrupt Climate Change working group. Yvan Romé's official profile focuses on climate–ice interactions, ocean circulation, D–O mechanisms, model development and comparison with palaeo-reconstructions. The project also names Lauren Gregoire at Leeds, and Andrew Keane and Camille Li at partner institutions.

Recent work gives a clear picture of the research lineage:

• Romé et al. (2025), “Simulated millennial-scale climate variability driven by a convection–advection oscillator.” The paper analyzes the mechanism behind oscillations in long HadCM3 simulations and relates it to stratification and salt exchange. Its own scope statement says the proposed mechanism may be simulation- and model-dependent, so the PhD's broader testing is genuinely open work.
• Romé et al. (2022), “Millennial-scale climate oscillations triggered by deglacial meltwater discharge.” This study reports Last Glacial Maximum simulations in which oscillation depended nonlinearly on the location and amount of freshwater forcing and on the background climate. It supplies the experimental foundation for asking which physical ingredients are essential.
• Snoll et al. (2025), “Competing effects of sea ice change control the pace and amplitude…” This team paper investigates sea-ice effects on millennial variability, reinforcing that ocean–ice feedbacks may modulate rather than uniquely determine the oscillation.
• Two independent 2024 studies offer useful methodological counterpoints: one challenges statistical AMOC warning claims because of proxy-coverage artefacts; another develops a physics-based indicator in a modelled tipping event.

These sources show active publication and an interdisciplinary network. They do not specify how often co-supervisors meet, who owns particular thesis chapters, whether code and model configurations are already prepared, or whether core data and high-performance computing time are allocated.

Funding, Application and Living Cost

YES-DTN says it offers approximately 26 studentships each year across Leeds and York projects. Its published award comprises a tax-free stipend at UKRI rates for 3.5 years, university fees, and support for research costs, fieldwork, training and conference travel. The FAQ says fees are covered irrespective of the home/international difference and that up to 30% of awards can go to international candidates. A three-month industry placement can extend the funded period to 3.75 years. The FAQ also says relocation, visa and health-surcharge costs are not covered. UKRI's current 2026/27 minimum stipend is £21,805; the 2027/28 rate for this entry year is not yet confirmed.

The University of Leeds' 2026/27 living-cost guide estimates an average £320 per week or £16,640 for 52 weeks, excluding tuition. That is below the current UKRI annual minimum by about £5,165 before individual circumstances, any additional costs, future inflation or changes in stipend. It is a comparison of published averages, not a guaranteed personal surplus; accommodation is the main source of variation.

Deadline and steps: the project description says Wednesday 13 January 2027, while official network instructions say applications close 13:00 GMT, 12 January 2027. The official procedure requires a Leeds Online Application and a separate YES-DTN application form, with a Leeds student ID entered on the second form. Submit the first form at least two days early, and verify the exact deadline on the live network application page before final submission.

Applicant Fit and Preparation

The project page lists a broad relevant-degree range, including applied mathematics, atmospheric and earth-system science, computer science, environmental science, geography, geology, geophysics, mathematics, meteorology, oceanography, physics and related areas. It does not state a formal minimum degree classification or list individual programming packages as mandatory.

The work is likely to reward applicants who can demonstrate quantitative reasoning and comfort crossing disciplinary boundaries. Useful preparation would include examples of numerical modelling, statistical analysis, physical ocean/climate processes, palaeoclimate records, or reproducible scientific computing. These are preparation signals inferred from the advertised tasks, not additional eligibility rules. Candidates should clarify whether the project expects prior climate-model experience or provides it, which languages and workflows are used, and what level of sediment-core data handling will be needed.

Applicants whose main interest is observational oceanography, policy, or statistical prediction alone should ask how the three objectives can be balanced; the advert commits the student to all three and foregrounds physical-model development. Likewise, applicants should avoid pitching a definite modern collapse date. The current evidence base does not support treating the PhD's early-warning task as a settled forecast exercise.

Next Actions and Questions to Verify

Before applying

• Ask Dr Ivanovic which model configuration, computing resources and training are available at the start, and how much of the thesis is expected to use modelling versus palaeo-record analysis.
• Confirm the exact 12 January 2027, 13:00 GMT cut-off and which two application forms/document requirements apply to this project.
• Ask how the international award cap is allocated, whether the 2027/28 stipend rate is set, and whether all overseas fees are paid directly under the award.
• Clarify whether the optional three-month placement is realistically available for this project and how it affects thesis scope and funded duration.

At interview

• Ask which competing D–O mechanisms are considered the most important comparisons, how model dependence will be tested, and what constitutes a successful validation against sediment-core records.
• Ask how the team will handle uncertainty in proxy chronology, coverage and early-warning false positives.

Before accepting an offer

• Obtain written confirmation of the award amount and duration, fee status and coverage, research-cost support, registration period, supervision arrangements, computing allocation and expected start date.

Evidence and Sources

Project and application

• YES-27-A-12 project description
• 2027 application instructions
• YES-DTN studentship funding
• YES-DTN funding and international-student FAQ

School and doctoral environment

• University of Leeds research degrees
• School research themes
• Physical Climate and Cryosphere cluster
• QS Earth & Marine Sciences ranking 2026
• University of Leeds address

Supervisors and recent research

• Dr Ruza Ivanovic profile
• Yvan Romé profile
• Romé et al. (2025), convection–advection oscillator
• Romé et al. (2022), meltwater-triggered oscillations
• Snoll et al. (2025), sea-ice controls
• Chen and Tung (2024), AMOC proxy warning-signal critique
• van Westen et al. (2024), physics-based AMOC warning signal
• Annual review of AMOC collapse probability

Funding, costs and careers

• UKRI stipend and fee levels
• Leeds 2026/27 living-cost estimate
• UK climate-scientist career profile
• Met Office climate-modelling data-science role