Home

Senior Research Fellow in Machine Learning for Weather and Climate Science at University College London. Theme lead for UCL’s Met Office Academic Partnership. My research interests focus on the intersection of data-driven approaches with high performance computing techniques in the area of atmospheric and ocean modelling.

Education

• Ph.D in Applied and Computational Mathematics, University College Dublin, Oct. 2017 - Jan. 2021
• B.Sc. in Theoretical Physics, University College Dublin, Sep. 2013 - May. 2017

Work experience

• Senior Research Fellow in Machine Learning for Weather and Climate, AI Centre, Department of Computer Sciences, University College London, Oct. 2023 - present
  • Acting deputy group leader of the Sustainability and Machine Learning Group.
  • Oversee and manage ongoing research projects.
  • Collaborate with various international research groups.
  • Co-supervise PhD students within the Departments of Statistical Sciences and Computer Science.
  • Lead the “Tools, Systems and Practices” Met Office Academic Partnership (MOAP) working group.
  • Supervise BSc and MSc students.
  • Teach as a guest lecturer for MSc modules.
• Senior Research Software Engineer/Data Scientist, Centre of Advanced Research Computing (ARC), University College London, Jan. 2023 - Oct. 2023
  • Managed ARC’s involvement in the UCL Centre of Digital Innovation (CDI). The UCL CDI is a collaborative initiative between UCL and AWS. Performed technical needs assessments and developed data processing pipelines for start-ups within the CDI Impact Accelerator programme. Worked with and supported the PhD funded researchers.
  • Led a project involving the optimisation of a cosmological code designed to run on peta-scale machines. The optimisation of the codebase involved implementing architecture aware mapping of distributed memory computations.
  • Managed assistant research software engineers.
• Research Fellow, Department of Statistical Sciences, University College London, Feb. 2021 - Jan. 2023
  • Developed an open source data assimilation software package which implements particle filtering techniques and has applications in atmospheric and tsunami modelling (Julia).
  • Collaborated with a UK Met Office partner to improve the representation of high resolution information within the Unified Model for climate predictions.
  • Partnered with AON (reinsurance) to develop real-time approaches for providing historical context to severe weather events.
  • Developed the numerical software on a project funded by AON (reinsurance) titled: “Production-level Cascadia tsunami hazard model incorporating far-field sources”.
  • Co-supervised a student’s MSc project focusing on developing machine learning approaches to predict atmospheric indices from real observations.
  • Associate Member of the Centre of Advanced Research Computing (ARC)
• Postgraduate Research Scholar, School of Mathematics and Statistics, University College Dublin, Oct. 2017 - Jan. 2021
  • Lead developer of a tsunami modelling software package optimised to utilise the latest high performance computing architectures
  • Created novel automated techniques which predict local variability of tsunami impact using transfer functions and machine learning methods.
  • Coupled a fast tsunami solver with a statistical emulator to provide tsunami warnings with associated uncertainties.
  • Vice-President of the SIAM-IMA Dublin Area Student Chapter
• Research Scholar, Department of Numerical Modelling, Paul Scherrer Institute (ETH Zurich), May. 2017 - Aug. 2017
  • Modelling of fluid flows in nuclear reactors with an emphasis on implementing particle tracking techniques to track radionuclides
  • Supervisor: Professor Bojan Niceno
• Research Scholar, Mathematics Applications Consortium for Science and Technology, University of Limerick, May. 2016 - Aug. 2016
  • Developed a mathematical model for predicting influential spreaders in complex networks and tested its performance by developing Python code and using idealised Erdos-R'enyi graphs
  • Supervisor: Professor James Gleeson

Academic Reviewer

Reviewed numerous submissions to the following journals

• Journal of Applied Mathematical Modelling; the European Journal of Mechanics /B Fluids; Wave Modelling; Estuarine, Coastal and Shelf Science; Applied Mathematics and Computing; Hydroinformatics; Natural Hazards and Earth System Sciences; International Journal for Uncertainty Quantification

Academic Supervisor

Academic co-supervisor for a student’s MSc project (UCL Machine Learning) (present)

• “Embedding a machine learning model within a global atmospheric model”

Academic co-supervisor for a student’s MSc project (UCL Scientific and Data Intensive Computing Masters) (May. 2021 - Sep. 2021)

• The student’s project was a collaboration with the UK Met Office and it focused on developing machine and statistical learning approaches for predicting atmospheric indices

Academic co-supervisor for a student’s (ENS Paris-Saclay) undergraduate internship (Apr. 2019 - Aug. 2019)

• The student’s work focused on developing ML techniques for capturing localised tsunami responses

Workshops & Conferences

Co-organised a mini-symposium at SIAM GS, Bergen 2023

• Fusing of machine learning with weather/climate modelling

Delivered a workshop at IIT Delhi, 2023

• Computing methods for Validation, Verification and Uncertainty Quantification

Awards & Scholarships

• University College Dublin Entrance Scholar (UCD, 2013)
• 2nd Year Academic Scholar (UCD, 2014)
• 3rd Year Academic Scholar (UCD, 2015)
• Science and Engineering Summer Research Scholarship (UL40) (University of Limerick, 2016)
• Summer Research Scholarship at the Paul Scherrer Institute (PSI, 2017)
• Government of Ireland Postgraduate Research Scholarship (2017)
• Bertram Broberg Memorial Medal (2021) - Awarded to the best PhD thesis in Mechanics across the Schools of Engineering, Mathematics & Statistics and Physics.
• ABSW Media Fellowship (2024)

Publications

Uncertainty Quantification of Pre-Trained and Fine-Tuned Surrogate Models using Conformal Prediction

Gopakumar, V., et al.: Uncertainty Quantification of Pre-Trained and Fine-Tuned Surrogate Models using Conformal Prediction, 2024.

Valid Error Bars for Neural Weather Models using Conformal Prediction

Gopakumar, V., et al.: Valid Error Bars for Neural Weather Models using Conformal Prediction, 2024.

Embedding machine-learnt sub-grid variability improves climate model biases

Giles , D., et al.: Embedding machine-learnt sub-grid variability improves climate model biases, 2024.

A collaborative hackathon to investigate climate change and extreme weather impacts in justice and insurance settings

Macholl, J.D., et al.: A collaborative hackathon to investigate climate change and extreme weather impacts in justice and insurance settings, 2024.

Transatlantic Data Science Academy Project. Phase 1: Scoping and Shaping

Dance , S., et al.: Transatlantic Data Science Academy Project. Phase 1: Scoping and Shaping, 2024.

Scalable Data Assimilation with Message Passing

Key, O., et al.: Scalable Data Assimilation with Message Passing, 2024.

ParticleDA.jl v.1.0: a distributed particle-filtering data assimilation package

Giles, D., Graham, M. M., Giordano, M., Koskela, T., Beskos, A., and Guillas, S.: ParticleDA.jl v.1.0: a distributed particle-filtering data assimilation package, Geosci. Model Dev., 17, 2427–2445, 2024.

Time-dependent influence metric for cascade dynamics on networks

Gleeson, J., Cassidy, A., Giles, D. and Faqeeh, A.: Time-dependent influence metric for cascade dynamics on networks, 2024.

Meteotsunamis and other anomalous “tidal surge” events in western Europe in summer 2022

E. Renzi, C. Bergin, T. Kokina, D. S. Pelaez-Zapata, D. Giles and F. Dias: Meteotsunamis and other anomalous “tidal surge” events in Western Europe in Summer 2022. Physics of Fluids, 35, 4, 2023

Multilevel Bayesian Quadrature

Kaiyu, L., Giles, D., Karvonen, T., Guillas, S. and Briol, F.X.: Multilevel Bayesian Quadrature, Proceedings of The 26th International Conference on Artificial Intelligence and Statistics, 206, 1845--1868, 2023.

Automated approaches for capturing localized tsunami response—Application to the French coastlines.

Giles, D., Gailler, A., & Dias, F.: Automated approaches for capturing localized tsunami response—Application to the French coastlines. Journal of Geophysical Research: Oceans, 127, 2022

Faster Than Real Time Tsunami Warning with Associated Hazard Uncertainties

Giles, D., Gopinathan, D., Guillas, S., and Dias, F.: Faster Than Real Time Tsunami Warning with Associated Hazard Uncertainties, Front. Earth Sci., 8, 226, 2021.

Performance analysis of Volna-OP2 – massively parallel code for tsunami modelling

Giles, D., Kashdan, E., Salmanidou, M. D., Guillas, S., and Dias, F.: Performance analysis of Volna-OP2 – massively parallel code for tsunami modelling, Computers & Fluids, 209, 104649, 2020.

Modelling with Volna-OP2—Towards Tsunami Threat Reduction for the Irish Coastline

Giles, D., McConnell, B., and Dias, F.: Modelling with Volna-OP2—Towards Tsunami Threat Reduction for the Irish Coastline, Geosciences, 10(6), 226, 2020.

The Volna-OP2 tsunami Code (version 1.5)

Reguly, I. Z., Giles, D., Gopinathan, D., Quivy, L., Beck, J. H., Giles, M. B., Guillas, S., and Dias, F.: The VOLNA-OP2 tsunami code (version 1.5), Geosci. Model Dev., 11, 4621–4635, 2018.