Skip to main content

Computer Science

Photo of Dr Charlie Kirkwood

Dr Charlie Kirkwood

Postdoctoral Research Fellow


I’m a research fellow in the University of Exeter’s Institute for Data Science and AI (IDSAI), where I work to develop and apply data-scientific solutions for a range of pilot projects proposed by academics from across the University (you can read more about the IDSAI’s pilot project scheme here). If you have ideas for how machine learning could help you in your research, and/or would like to know more, please feel free to get in touch.

I have also been developing my own research programme in artificial intelligence for geological mapping, for which my article ‘Geological Mapping in the Age of Artificial Intelligence’ featured on the front cover of the Geological Society of London’s magazine (supported by my various technical papers dating back to 2016). I have given several invited lectures on this topic, including to EuroGeoSurveys which is a consortium of the national geological surveys of all European countries. You can watch a recording of my talk on the EuroGeoSurveys YouTube channel here. If you are interested in AI and geology, please free to get in touch about this too!

Prior to joining the IDSAI as a research fellow I obtained my PhD in Mathematics from the University of Exeter, where I was funded by the EPSRC and the Met Office. My thesis was entitled ‘Methods in machine learning for probabilistic modelling of environment, with applications in meteorology and geology’ for which I worked at the intersection of ensemble weather forecasting and probabilistic AI, and applied the same concepts to environmental modelling more broadly. Two of my PhD publications are cited in the Met Office’s Data Science Framework 2022-2027. Outputs from my PhD also featured on the Society of Industrial and Applied Mathematics (SIAM) news blog and in a University of Exeter press release.

Before my PhD, I originally graduated in 2012 with an MESci in Exploration and Resource Geology from Cardiff University and started out working in the mineral exploration industry, where geostatistical methods are a key part of decision making. I then moved into a NERC funded research position at the British Geological Survey where I developed more advanced statistical and machine learning methods for quantitative geological mapping (see publications, and website). Keen to pursue my quantitative interests beyond the boundaries of geology, I later took a job as a commercial data scientist at Walgreens Boots Alliance (parent company of Boots on the UK high street), where I developed machine learning methods to improve personalised targeting of offers to customers, and natural language processing systems for prioritising responses to customer feedback.

Grants and Awards:

2018 - Exeter University Researcher-Led Initiative grant, which I used to co-organise two courses - 'An introduction to the Git ecosystem for version control and code sharing' and 'An introduction to data analysis in Python', both kindly taught by Dr Chris Woods, Research Software Engineering Lead at the University of Bristol.

2018 - WBA Global Brands 'Makers' award, for advancing data science at Boots UK.

2017 - Elsevier Outstanding Reviewer award for my contributions to the Journal of Geochemical Exploration.

2016 - NVIDIA GPU grant, awarded a Titan X Pascal GPU to accelerate research into the use of neural networks & deep learning in geological mapping and mineral exploration (for example see poster here).

Back to top


Copyright Notice: Any articles made available for download are for personal use only. Any other use requires prior permission of the author and the copyright holder.

| 2023 | 2022 | 2021 | 2020 | 2018 | 2016 |


  • Kirkwood C. (2023) Methods in machine learning for probabilistic modelling of environment, with applications in meteorology and geology.




  • Kirkwood C, Economou T, Pugeault N. (2020) Bayesian deep learning for mapping via auxiliary information: a new era for geostatistics?. [PDF]
  • Kirkwood C. (2020) Deep covariate-learning: optimising information extraction from terrain texture for geostatistical modelling applications. [PDF]



Back to top