Overview

I am a Bayesian statistician and Professor of Uncertainty Quantification in the Land Environment Economics and Policy Institute. I suppose I could also be called "Professor of Bayesian Statistics" or "Professor of Data Science", but you have to settle on one title in the end. I have always been interested in decision making under uncertainty, with a particular focus on what uncertainty means (to analyst and decision maker), how to quantify it, how to use it to guide decision makers and when it is ethical to do so. A feature in the vast majority of my work has been the presence of one or more expensive computer simulators (or "models") of the aspects of reality we are uncertain about, and most of my application work has been with environmental models (e.g. atmosphere, ocean, land ice, coupled GCMs, land surface, crop growth). I am a subjectivist, so I think uncertainty is nothing more than a property of individuals, probability is one possible calculus for its quantification and communication, and “true randomness” may or may not exist, but it doesn’t really matter. I am also a Dad, a chess player, a cyclist, a (beginner) sea-kayaker and I play fantasy football (without any statistical modelling at all).

Qualifications

I have an MMATH and a PhD in Statistics from Durham University. My 2010 PhD thesis was titled: “Policy making using computer simulators for complex physical systems: Bayesian decision support for the development of adaptive strategies.”

Career

• 2010-2013: Post-Doctoral Research Associate at Durham University.
• 2013-2016: EPSRC fellow and Lecturer: Department of Mathematics, University of Exeter.
• 2016 – 2020: Senior Lecturer and Director of Impact, Department of Mathematics, University of Exeter. 
• 2020 – 2023: Associate Professor of Bayesian Statistics, Director of Impact, Department of Mathematics and Statistics, University of Exeter.
• 2023 – Present: Professor of Uncertainty Quantification, Land Environment Economics and Policy Institute, Department of Economics, University of Exeter. 

Highlights:

• International Society for Bayesian Analysis Savage Award (for outstanding methodological PhD thesis) 2011
• EPSRC Fellow 2013-2016.
• International Society for Bayesian Analysis Lindley Prize 2014
• Alan Turing Institute Fellow 2019-2023

Back to top

Research

Research interests

The study and implementation of methods for quantifying uncertainty about the real-world using computer models (referred to in the literature as "Uncertainty Quantification (UQ)"), is my principle methodological interest, and I have mainly worked in environmental applications of these ideas. The “emulation” of computer models, so that they can be interrogated in real time, calibrated to observations and even coupled within efficient decision support tools is a core area of UQ. I have contributed to this area, particularly for environmental models, with extensions to Gaussian processes for non-stationarity, spatial output (via dimension reduction and kernel methods), and deep and linked Gaussian processes for coupling. I have made contributions to the topic of calibrating models to observations whilst managing structural error, in particular, using History matching, with extensions particularly for spatio-temporal data. I have developed and demonstrated these techniques in application to “climate model tuning”. 

I am interested in the development of digital twins, particularly the embedding of UQ to deliver real-time interrogation, coupling and calibration of models to data streams for digital twins to be used in decision support. I’m also interested in Ethical AI from a foundational perspective: how can the uncertainty used or reported by AI systems be understood and ethically used by decision makers, particularly where multiple competing models/frameworks can be deployed and may give different responses. Recently I have been considering Land Use Change problems, and in particular for delivering Net Zero, within the context of the all of the above (UQ, digital twinning and ethical AI). 

Research projects

PI for ADD-TREES an EPSRC AI for Net Zero project with full title AI-elevated Decision-support via Digital Twins for Restoring and Enhancing Ecosystem Services. https://netzeroplus.ac.uk/add-trees-project/

AI lead for Net Zero Plus, A Greenhouse Gas Removal Demonstrator focused on tree planting. https://netzeroplus.ac.uk

Back to top

Publications

Back to top

Supervision / Group

Postdoctoral researchers

Postgraduate researchers

Back to top