Back to the Future: Ice sheet collapse, ocean circulation slowdown and abrupt climate change

Background

The Atlantic Meridional Overturning circulation (AMOC) – a large-scale system of ocean currents – is critical for keeping North American and European climates mild. A disruption to AMOC would have dramatic consequences, but models disagree on how AMOC will respond to future climate warming and Greenland melting. Modern observations are too short and do not capture large enough changes to sufficiently test or calibrate modelled AMOC sensitivity to 21st Century climate change. The best way to test models is to turn to the last episode of AMOC disruption, 8,200 years ago, when accelerated ice sheet melt caused AMOC slowdown and a century-long 1-3ºC Northern Hemisphere cooling. Touted as the “Goldilock’s” event for benchmarking future AMOC change, this so-called ‘8.2 kyr event’ has quantified forcing and abundant palaeoclimatic records for testing model responses. We can use this event to inform future climate projections thanks to new techniques in statistical Uncertainty Quantification.

PhD opportunity

Using advanced uncertainty quantification, this project will combine complex coupled climate-ice sheet models with geological records of Earth’s most recent ice sheet collapse and reorganisation of Atlantic Ocean circulation 8,200 years ago to provide ‘out of sample’ constraints on the sensitivity of ocean circulation to future Greenland melt, improving confidence in future climate projections.

Efficient experimental design (eg. multi-wave Latin Hypercube Sampling) will be used to run ensembles of complex Earth Systems Models (incl. UKESM, the UK flagship model with interactive ice sheets) for the 8.2 kyr event, the present day and future, varying uncertain model parameters and ice sheet melt. Advanced Gaussian process emulators (statistical models used in artificial intelligence/machine learning) will be applied to learn the relationships between uncertain model inputs and past/future AMOC change, to: (i) optimally exploit information from different models; (ii) understand the link between past and future changes. Implausibility metrics will rule out unrealistic models within uncertainty, producing a probability distribution of plausible future AMOC change. The project could apply the same techniques to constrain processes of ice sheet collapse, but with ice sheet model ensembles and reconstructions.

These tools will be used to answer exciting research questions chosen by the postgraduate researcher with support from the supervisory team, eg:

– What aspects of future climate projections can be constrained with observations of the 8.2 kyr event (amplitude, timing, pattern of change)?

– What additional observations might we need for the 8.2 kyr event to constrain future AMOC change?

– Can simulating the final collapse of the North American ice sheet help test and improve future Greenland projections?

The project may also investigate the usefulness of the 8.2 kyr event as a real-world ‘storyline’ to help decision-makers prevent or adapt to future climate change.