I work at ECMWF as a research scientist within the ocean modelling team, developing and evaluation data-driven ocean models. My work focuses on using and adapting ANEMOI to train predictive models of the ocean, using machine learning. This includes assessing different architecture options, data sources and resolution, and training methodologies.
- Data-driven ocean modelling
- Assessment of machine learning models
- Impact of ocean models on earth system predictions
I studied mathematics at Oxford university, before joining the UK Met Office, working in the 'ocean forecasting research and development' team. I worked developing NEMO to include coastal modelling and storm surge modelling capability, building and assessing configurations for operational use.
In 2018 I began a PhD with Cambridge University and the British Antarctic Survey on data-driven ocean modelling. This focused on developing data-driven emulators of an idealised channel configuration of MITgcm. As well as developing and assessing data-driven emulators, my research looked at the stability of these data-driven approaches, and model interpretability -- methods to understand what has been learnt by the machine learning algorithms and how well this represents the known dynamics of the system.
Through my career I've been on the steering and organising committee of many conferences, including convening the popular Machine Learning for Ocean Modelling session at EGU, and co-organising the Liege Colloquium on machine learning and data analysis in oceanography. I have also taken part in various panels and outreach events.
- ARIA VERIFY external advisory board member (2025 - present)
- Cambridge Centre for Data Driven Discovery steering committee ECR member (2019-2022)
- NEMO steering committee member (2010-2014)
ecmwf.int