About the Project
Background
Large hail is a rare but impactful weather hazard, capable of causing extensive damage to homes, vehicles, crops and infrastructure. European hail observations come from a diverse range of sources including weather radar, hail pads, insurance claims data and citizen science, differing greatly in spatial coverage, quality and record length. This project will develop new statistical methods to combine these diverse data sources, producing a more robust picture of the distribution of large hail…
events and reducing uncertainty in our understanding of changing atmospheric perils under a warming climate.
A central aim will be to quantify how the risk of severe hail may evolve over the coming years as the climate warms. The student will develop methods to account for missing or sparse observations, and distinguish changing risk from changing observation practices. This project will benefit from partnership with Liberty Mutual, a global insurer with extensive expertise in atmospheric perils.
PhD project
A warming climate may be associated with extreme atmospheric perils, one being large hail events, a rare but highly damaging weather hazard. This project will quantify how the risk of severe hail may evolve over the coming years, utilising all available data on this peril effectively. The project will draw on statistical techniques from extreme value analysis, where theoretically-motivated approaches allow for extrapolation and estimation in a distribution’s tails.
Model-building here requires tackling challenging data integration problems, leading to missing data problems. Data may be measured at different spatial locations or resolutions, with structural gaps occurring when they are combined. There may be varying levels of quality associated with each data input. Data may also be unrepresentative due to some locations being easier to access than others. To address these, the project will investigate using Bayesian methods to model extreme events that simultaneously deal with missing values within a unified framework. Statistical or machine learning methods will also be explored to deal with Structured Missingness arising from integrating data sources with incomplete overlap.
The primary supervisor (Prof Mitra) is an expert in missing data methods with the co-supervisor (Dr Simpson) an expert in extreme value modelling, external partners are Dr Rhodes and Dr Hernandez at Liberty Mutual.
Liberty’s involvement grounds the project in real-world relevance facilitating a direct route to practical impact. As a global insurer with substantial exposure to hail risk, partnership with Liberty will provide an industry perspective on how hail risk is currently managed, and a view on the relevance of this research within the exposure management and catastrophe modelling space. Guidance provided will ensure the research aligns with industry-relevant questions. Regular contact with researchers at Liberty also gives the student the opportunity to build a professional network.
Applicant Profile
Students with a strong background in mathematics and statistics, who have an interest in developing new statistical methodology with applications in climate science. Experience with coding in R will also be useful.
Funding Notes
This project is part of the UNRISK CDT, which offers 15-18 fully-funded NERC studentships, covering full university tuition fees; a personal stipend at standard UKRI rates; £6000 individual research and training costs; £5000 (per student) of cohort-level training; and a ‘Flexible Fund’ for special projects.
International applicants will need to cover costs related to applying for a student visa and the international health surcharge (IHS)
Applications are open to UK and international applicants. The number of awards for international applicants is limited by UKRI rules.
More information is available on the UNRISK website: View Website
References
Battaglioli, F. et al. “Modeled multidecadal trends of lightning and (very) large hail in Europe and North America (1950–2021).” Journal of Applied Meteorology and Climatology, 62(11), 1627-1653 (2023). https://doi.org/10.1175/JAMC-D-22-0195.1
Mitra, R., et al. “Learning from data with structured missingness.” Nature Machine Intelligence, 5, 13-23 (2023). https://doi.org/10.1038/s42256-022-00596-z
Simpson, E. S., and Northrop, P. J. “Accounting for missing data when modelling block maxima.” Environmetrics, 37(2), e70075 (2026). https://doi.org/10.1002/env.70075