In June 2025, a team of STOR-i PhD students Kristina Bratkova, Kajal Dodhia, Matt Speers and Wanchen Yue, and School of Mathematical Sciences PhD student Aiden Farrell, took part in the Extreme Value Analysis (EVA) Conference (2025) Data Challenge. The aim of the challenge was to estimate various probabilities of extreme precipitation using an ensemble of rainfall simulations. The difficulty of the challenge lay in the variety of probabilities the teams needed to estimate, and that these quantities generally exceeded the range of the observed data. The teams were scored on the accuracy of their estimates and confidence intervals.

The top 5 teams were invited to present their work at the Conference hosted by the University of North Carolina at Chapel Hill, United States. Team LancasterConditional was awarded second place, applying a methodology based on a conditional extreme value model framework with extensions to account for seasonality and temporal dependence in the margins, as well as a flexible copula to capture the complex multivariate spatial dependence.

Two team members provided their perspective on the experience. Kajal, a first-year STOR-i PhD student, said: "Taking part in the EVA Data Challenge furthered my knowledge of extreme value theory whilst learning a lot from my team, including how to interpret data from different perspectives, apply various statistical techniques, and collaborate effectively under time pressure.”

Second-year STOR-i PhD student, Wanchen, noted: "It was a rewarding opportunity for us to take our research beyond the academic setting and apply it to a real-world challenge. The diversity of ideas and approaches across the teams at the conference made it a genuinely inspiring event to be a part of. It was both a challenge and a joy!"

Well done team LancasterConditional!