Reducing Uncertainty in the Response of the Earth System to Anthropogenic CO2 Emissions with Statistical Emulation of Land Surface Models. PhD in Mathematics (NERC GW4+ DTP)

Lead Supervisor
Dr Oliver Stoner, Department of Mathematics, College of Engineering, Mathematics and Physical Sciences, University of Exeter

Additional Supervisors
Dr Anna Harper, Department of Mathematics, College of Engineering, Mathematics and Physical Sciences, University of Exeter
Prof Peter Challenor, Department of Mathematics, College of Engineering, Mathematics and Physical Sciences, University of Exeter

Location: University of Exeter, Streatham Campus, Exeter, EX4 4QJ

This project is one of a number that are in competition for funding from the NERC GW4+ Doctoral Training Partnership (GW4+ DTP). The GW4+ DTP consists of the GW4 Alliance of research-intensive universities: the University of Bath, University of Bristol, Cardiff University and the University of Exeter plus five unique and prestigious Research Organisation partners: British Antarctic Survey, British Geological Survey, Centre for Ecology & Hydrology, the Natural History Museum and Plymouth Marine Laboratory. The partnership aims to provide a broad training in the Earth, Environmental and Life sciences, designed to train tomorrow’s leaders in scientific research, business, technology and policy-making. For further details about the programme please see http://nercgw4plus.ac.uk/

For eligible successful applicants, the studentships comprises:

- A stipend for 3.5 years (currently £15,009 p.a. for 2019/20) in line with UK Research and Innovation rates
- Payment of university tuition fees;
- A research budget of £11,000 for an international conference, lab, field and research expenses;
- A training budget of £3,250 for specialist training courses and expenses.
- Travel and accommodation is covered for all compulsory DTP cohort events
- No course fees for courses run by the DTP

We are currently advertising projects for a total of 10 studentships at the University of Exeter.

Project Background

The land surface removes about one-third of human emissions of carbon dioxide (CO2) from the atmosphere each year. This carbon is stored in vegetation and soils. The future of the land surface response to climate change is highly uncertain, and some models predict that the land could switch from a sink to a source of CO2 during this century. This project focuses on JULES, a process-based computer model that represents the exchange of CO2, heat, moisture, and momentum between the land surface and the atmosphere in the UK’s climate model. JULES has more than 60 parameters which affect processes such as photosynthesis, respiration, evaporation, and plant growth and mortality. There is an urgent need for rapid improvements and better quantification of uncertainty in models such as JULES because of the importance of climate change and the need for science to support commitments to keep climate change to below 2°C.

Project Aims and Methods

An efficient method for calibrating numerical models like JULES is to employ a statistical (or machine learning) model which emulates the relationship between the input parameters and the mathematical processes, in this case of the land surface. Using this emulator, it is then possible to systematically study the effect of previously untried parameter combinations on the model’s ability to reproduce observed data, without having to run the original model too many times (which would in the case of JULES be computationally expensive). The aims of this project are therefore to:

Develop an emulator for JULES, to quantify the effect of varying model parameters on the land surface processes.
Use the emulator to calibrate JULES so that it can better reproduce historical observations. Specifically, the project will focus on seasonal and interannual variations in photosynthesis in dryland ecosystems. Despite having low rainfall and relatively sparse vegetation, drylands could store up to 25% of the land surface carbon. The student will compare JULES to available remote sensing data (for example leaf area index as shown in the image) to improve the model’s representation of the carbon cycle in these ecosystems.
Use the improved JULES to quantify the change in carbon storage in dryland ecosystems under future climate change. This will involve running the improved JULES with several different climates.

The supervisors are open-minded to adapting the project to suit the candidate’s research interests, e.g. to:

- explore other approaches to emulation (e.g. machine learning);
- place greater emphasis on the methodological development of the emulator or place greater emphasis on the use of the emulator to improve climate predictions.