Combining models and uncertainties to support flood risk assessment and mitigation strategies

Background

Surface water flooding events like those in Valencia in 2024 are becoming increasingly common in a wetter climate and with increasing urbanisation, affecting households and communities both through the direct experience of flooding and through the impact of rising insurance prices. Forecasting, understanding, and mitigating the impacts of surface water flooding are key climate adaptation challenges for the UK. The new National Flood Risk Assessment (NaFRA II, 2025) provides a set of up-to-date models, which we will build upon and use to provide evidence to support decisions about property-level flood resilience measures. This project combines mathematical/computational approaches (blending information from multiple models) with decision support (using the output to inform strategic flood resilience decisions).

PhD opportunity

This project is a CASE partnership with Flood Re, a UK reinsurer set up in partnership with the government to help households understand and reduce high levels of flood risk. Flood Re will operate until 2039, enabling a transition to home insurance prices that fully reflect flood risk. In 2025 the new National Flood Risk Assessment (NaFRA II) will be released, an open-source dataset covering flood risk from rivers, surface water and coastal flooding in unprecedented detail. However, the NaFRA Risk of Surface Water Flooding map only covers England and Wales and due to the use of detailed hydrological models, is slow to update. This project will address both issues by developing new methods (possibly machine learning based) for mapping surface water flood risk and uncertainty.
We will develop an appropriate methodology to bring together multiple models, taking into account associated uncertainties including climate variability and model error. This methodology will be applicable beyond the scope of flood. Next, we will consider how the unified risk mapping framework can be used in decision-making on annual to decadal timescales. For example, it might inform suitable strategies for prioritising investment in flood defences or the installation of property-level flood resilience measures. We will contrast the UK context with the Netherlands, where flood risks and uncertainty are managed through a different governance framework. Supervisor Erica Thompson is expert in using models to inform decisions; co-supervisor Arthur Petersen has extensive experience in climate risk assessment including in the Netherlands.
The student will develop an operational familiarity with flood risk, model evaluation and decision-making, and will spend time at the Flood Re office in London. The student would then be well-placed either to continue an academic career at the intersection of maths, climate risk and public policy, or to move into catastrophe modelling, (re)insurance or policy.

Other information

Applicant profile: Students with strong quantitative and computational skills who want to develop a nuanced understanding of risk and uncertainty, and gain experience in using mathematical models for pragmatic decision support in public and private sectors.

  • https://www.gov.uk/guidance/updates-to-national-flood-and-coastal-erosion-risk-information
  • https://www.floodre.co.uk/