skip to primary navigationskip to content
 

C402: New modelling tools for addressing computationally challenging climate science problems involving the ocean carbon cycle (Priority project with CASE partner) (Lead Supervisor: Emily Shuckburgh, British Antarctic Survey)

Supervisors: Emily Shuckburgh (British Antarctic Survey), Peter Haynes (DAMTP), Dan Jones (British Antarctic Survey) and CASE partner co-supervisor

Importance of the area of research:

Addressing some of the key questions in climate science from a modelling perspective is computationally challenging due to the need to run long simulations of models with fine spatial or temporal resolution and/or the inclusion of complex physical or biogeochemical processes.

One set of such questions involves the ocean carbon cycle. A critical gap in our understanding of past, present and future climate surrounds the response of the ocean carbon cycle to changes in forcing from the atmosphere. Looking to the past, temperature changes between ice ages and interglacial periods cannot be explained by orbitally driven changes in solar insolation alone; it is thought that feedbacks involving the ocean carbon cycle, in particular in the Southern Ocean must have been significant. Current estimates indicate that the net uptake of carbon dioxide by the Southern Ocean exceeds the emissions of the United States. Looking to the future, climate change could be strongly influenced by feedbacks influencing this uptake, but lack of knowledge of the sensitivity of the ocean carbon cycle to forcing is a source of significant uncertainty in future projections with important policy implications.

Project summary:

New, computationally-efficient modelling techniques are needed to enable some of the critical questions related to the ocean carbon cycle to become tractable. Ocean carbon cycle modelling is computationally intensive: it takes several thousand years for the deep-ocean to come into equilibrium with surface fluxes; many of the critical dynamical processes that must be resolved by ocean models to represent the relevant ocean circulation are small-scale (of the order 10km in the Southern Ocean), and the more realistic biogeochemical models include a detailed description of lower-trophic level marine ecosystems that can double or even triple the model's computational cost. The project will develop and test new modelling methodologies and use them to explore the sensitivity of the carbon cycle in the Southern Ocean to atmospheric forcing.

What the student will do:

This studentship will explore approaches to generating computationally-efficient ocean carbon cycle models. This may involve generating a reduced model for the ocean dynamics and coupling a detailed biogeochemical model, generating a reduced model of the biogeochemistry to couple to a high-resolution ocean model, or a combination of these. Approaches using inference to generate model reductions based on initial conditions, inputs, trajectories of the states, and outputs of the full model (e.g. Peherstorfer and Willcox, 2016; Khatiwala et al, 2005) will be developed and compared against alternative coarse-graining approaches (e.g. Bardin et al, 2014). Bayesian approaches will be used to assess the veracity of the model output in the context of observational data (with uncertainties) from the present day and from proxy measurements from the last glacial maximum. The new modelling tools developed will then be deployed to explore climate feedbacks involving the Southern Ocean carbon cycle and the implications for future climate change.

Please contact the lead supervisor directly for further information relating to what the successful applicant will be expected to do, training to be provided, and any specific educational background requirements.

References:

Peherstorfer, B.,  and Willcox, K. (2016). Data-driven operator inference for nonintrusive projection-based model reduction. Computer Methods in Applied Mechanics and Engineering, 306, 196-215.

Khatiwala, S., Visbeck, M., & Cane, M. (2005). Accelerated simulation of passive tracers in ocean circulation models. Ocean Modelling, 9, 51-69.

Bardin , A., Primeau, F., and Lindsay, K. (2014). An offline implicit solver for simulating prebomb radiocarbon. Ocean Modelling, 73, 45-58

Follow this link to find out about applying for this project.

Other projects available from the Lead Supervisor can be viewed here.

Filed under: