• Project
  • Host
  • How to Apply

Overview

The Earth's climate system is already significantly warmer than it has been for thousands of years. Short of almost unimaginably radical measures or unforeseen events, it is likely to become substantially warmer than it has been for many millions of years before the end of the 21st Century. This presents a problem for climate and Earth system science, as the vast majority of observational data available for comparison with future modelling exercises comes from more recent geological eras that are colder than the present. Relatively little detailed data from warmer climates exists and very little Earth system modelling of such warm states has been conducted to verify whether models can successfully reproduce these warm climate states.

Much warmer, 'greenhouse' climates, more closely analogous to expected future states existed in the Eocene epoch, around 35 to 55 million years ago2, but observational data from this period has hitherto been too sparse in time and space to determine the major features of the climate. Of particular importance is the pattern of deep ocean circulation currents that are the principal internal control on the climate system, and information on how they varied in time. Recent, much higher-resolution data from ocean sediment cores open the possibility of comparing the behaviour of the Earth system in these warm 'greenhouse' states with better understood variability in the more recent past3,4.

In the last few millions years climate has oscillated repeatedly between glacial and interglacial states on timescales of 10-100 thousand years,  with huge ice sheets advancing and retreating periodically over Northern continents, driven by small changes in solar insolation as a result of asymmetry in the Earth's orbit around the sun. Whether the same or similar patterns of variability would pertain in much warmer climate states without major ice sheets is currently unknown. On these timescales the principal internal control on Earth system dynamics is the ocean circulation and carbon cycle system, and possibly also the terrestrial carbon system of vegetation and soils. The appropriate tool to investigate this long-timescale variability is therefore a computational Earth system model that includes the nonlinear and three-dimensional dynamics and feedbacks involved in the ocean and land carbon cycles, but is computationally simple enough to run for hundreds of millenia.

 

The principal question will be the extent to which a range of plausible circulation patterns are susceptible to patterns of solar forcing perturbation relevant to insolation changes during orbital cycles. This will be assessed by running a range of perturbation experiments with the model, and ultimately by testing the stability of the system in long (100 kyr timescale), orbitally forced simulations.

Eocene surface ocean dissolved inogranic Carbon isotopic composition simulated by GENIE1

Methodology

The project will study the stability and variability of the Eocene climate and ocean circulation using the GENIE Earth system model of intermediate complexity (EMIC). Configurations of GENIE for the Eocene already exist1 (Fig 1.), and can easily be modified for the precise periods targeted in the project. Wind fields will be derived from the more complex GENIE-PLASIM Earth system model with detailed atmospheric dynamics. GENIE will also be configured to produce appropriate biogeochemical data for comparison with observational data from ocean cores, including carbon isotopes, oxygen concentration and other circulation and biogeochemical tracers.

Uncertainty and stability will be addressed using ensembles of simulations based on existing calibration exercises that have mapped out plausible regions of model parameter space3,5. Bayesian statistical methodology will be used to develop statistical emulators of the modelled ensembles and assess the uncertainty in modelled scenarios. 

Training and Skills

CENTA students will be provided with 45 days training from CENTA through their PhD which includes a 5-day residential and a 10-day work placement. In the first year, students will undertake training in general environmental science, research methods and core skills as a single cohort. Training in years 2 and 3 will progress from core skills to masterclasses specific to the project and overall scientific theme. 

The student will receive training in using the GENIE-1 Earth system modelling framework and PLASIM dynamical atmosphere model; emulation; and other numerical, statistical, or data processing methods required for the project. These are likely to include: experimental design for computer experiments, Bayesian calibration, R statistical software, and data visualisation.

Students will also receive training in communicating science to different audiences (peers, the public, school students, media and policy makers) using a variety of methods including presentations, social media, policy briefing notes, public events and interviews. 

Timeline

Year 1: Setting up and calibrating the Eocene configuration of GENIE with appropriate paleogeography and wind fields; devising a method for analysing stability to orbital perturbations.

Year 2: Running large ensembles of simulations to probe Earth system stability in the Eocene; comparing with observational data and other model studies; preparing first manuscript for publication.

Year 3:Running long-timescale simulations to test hypotheses concerning stability and variability in key periods and background states; presenting results at two international conferences; writing up results for thesis and further publishable papers.

Partners and collaboration (including CASE)

Professor Andy Ridgwell (Bristol and UC Riverside, California) developed the ocean biogeochemistry component of GENIE. The student will visit Bristol for help with the biogeochemical modelling.

Alex Dickson (Oxford University) is an expert on the geochemical data relevant to the project. The student will spend time in Oxford working with Dickson on data-model comparison.  

Further Details

Students should have an undergraduate degree in or closely related to physics, mathematics or geophysics, and an enthusiasm for numerical (computer) modelling. Experience in programming and/or applied statistics would be an advantage but is not essential.

Interested candidates are encouraged to contact Neil Edwards neil.edwards@open.ac.uk for more information.