An interdisciplinary research project combining statistical modelling and climate adaptation aims to improve predictions of extreme rainfall and support better climate adaptation decisions in Denmark and neighbouring regions.
Extreme rainfall events are becoming increasingly frequent in Denmark, causing flooding, overloaded sewer systems with wastewater overflows, and significant economic costs and disruption in cities, towns, and the countryside.
Over the next three years, Associate Professor Anders Stockmarr from DTU Compute and Professor Peter Steen Mikkelsen from DTU Sustain, together with their research teams, will develop new statistical models for predicting extreme rainfall events. The aim is to provide a more robust evidence base for future climate adaptation planning.
The project builds on a previous joint PhD project in collaboration with the Danish Meteorological Institute (DMI) that explored how advanced mathematical modelling can improve our understanding of extreme rainfall events in Denmark. Find the links to two of the papers in fact box below!
“We have already demonstrated that our approach can, in some situations, provide more accurate estimates than the models currently in use. We now want to take the next step and develop methods that can better account for the changes we expect because of climate change,” says Anders Stockmarr. And Peter Steen Mikkelsen adds:
“If we can reduce the uncertainty in the calculations that underpin climate adaptation decisions, policymakers and planners will have a stronger basis for action before committing to costly infrastructure investments.”
The project is one of six DTU Compute projects awarded funding in the latest round of the Villum Synergy programme.
“DTU Compute’s involvement in half of the Villum Synergy grants awarded in 2026 reflects our strong position in data-driven and AI-based research, and the significant research potential that comes from working across research domains. We greatly value the opportunities provided by the Villum Foundation to develop new research collaborations across disciplines,” says Head of Department, Professor Jan Madsen.
Weather patterns are changing and local conditions matter
When utility companies and municipalities design sewer systems and other infrastructures, their decisions are based on estimates of the magnitude of rainfall events that may occur in the future.
Traditionally, these estimates rely only on historical observations from rain gauges. However, factors such as ocean currents, wind patterns and other known drivers of weather may have an influence, as well as a range of random processes that are more difficult to describe. At the same time, local conditions are important. Landscape features, coastlines, and topography all influence where rainfall occurs and how intense it becomes.
“DTU Compute’s involvement in half of the Villum Synergy grants awarded in 2026 reflects our strong position in data-driven and AI-based research, and the significant research potential that comes from working across research domains,” says Jan Madsen Professor, Head of Department at DTU Compute
The researchers will combine meteorological data with advanced statistical methods capable of capturing both the spatial and temporal variation in extreme rainfall.
“Many existing models are fundamentally based on the assumption that the climate remains broadly stable over time, even though the weather one year is not the same as the year before. We know that both natural variation and long-term climate change are taking place. We therefore want to develop models that account for these developments, enabling more reliable planning and dimensioning of sewer systems, drainage infrastructure, and future climate-resilient infrastructure more generally,” Anders Stockmarr says.
A synergy between statisticians and climate researchers
The new grant will enable close collaboration between statisticians and researchers in climate and environmental science, bringing together advanced mathematical modelling and domain-specific knowledge of climate systems.
Mathematics can reveal patterns that climate specialists may not have anticipated. Conversely, climate science helps ensure that the models reflect real-world processes in meaningful ways.
“We can develop new mathematical tools, but it is through collaboration with domain experts that we can see how they perform in practice,” says Anders Stockmarr.
“At the same time, new models inspired by other research fields can provide new insights into the phenomena that impact climate-resilient water systems” says Peter Steen Mikkelsen.
An example from the previous PhD project concerned the role of proximity to the coast. Researchers initially expected coastal location to be a key factor in explaining extreme rainfall events. However, their analyses showed that this effect for lesser thresholds for extreme rainfall was captured through the model’s representation of local weather dependencies and spatial patterns, and the effect therefore appear to manifest for heavier rainfall events only.
More data provide a clearer view
A key part of the project is to make use of a much broader range of data sources than is common today. Current approaches do not fully exploit the combination of traditional rain gauges, which collect rainfall physically in a measuring container, modern weather-station data, and private sensor networks.
In addition to official rainfall measurements, the researchers will investigate how data from local weather stations and other privately operated monitoring networks can be incorporated into their analyses. These may include rain gauges owned by private citizens, businesses, or farmers. Such data can be used to validate and improve the mathematical models.
“The more reliable observations we are able to include, the better our understanding of how rainfall varies at a local level. This can provide a much more detailed picture of where and how extreme events arise,” says Anders Stockmarr.
The text continues below the photo.
Models with applications far beyond Denmark
The researchers already collaborate with DMI on climate data and expect the results to contribute in the longer term to future climate services and climate scenario development.
Their ambitions extend well beyond Denmark.
“We hope that, within three years, we will have models that significantly improve our ability to support future climate adaptation decisions. We also hope to demonstrate that the methodology can be applied beyond Denmark,” says Anders Stockmarr.
In the longer term, the ambition is to scale the approach to larger parts of Northern Europe, where many of the same weather systems affect countries across national borders. This will require models that can be adapted to local conditions.
“Denmark is a relatively flat country with limited variation in elevation and comparatively modest geographical diversity. In countries such as the United Kingdom, for example, the geology, hydrology, and topography are very different, with much greater variation across the landscape. The objective is therefore not simply to transfer a Danish model directly, but to develop a methodology that can be adapted to the specific conditions of each country,” says Peter Steen Mikkelsen.
In this way, the researchers hope not only to contribute to improved climate adaptation in Denmark, but also to establish a scientific foundation that can support efforts to address future climate challenges more broadly.
Facts
Villum Synergy Programme
The grant of DKK 4,440,117 awarded to the project STEM-NEXT – Spatio-Temporal Extreme Rainfall Modelling for Next-Generation Hazard Assessment will fund, among other things, two postdoctoral researchers, one at each department. The project will begin on 1 January 2027.
The project builds on a previous joint PhD project in collaboration with the Danish Meteorological Institute (DMI) that explored how advanced mathematical modelling can improve our understanding of extreme rainfall events in Denmark. Read more in the two papers:
- Antoniadou N , Pedersen JW , Stockmarr A, Sørup HJD, Schmith T, Mikkelsen PS (2026):A Bayesian spatial framework for modeling sub-hourly to daily extreme precipitation in Denmark using SPDE with INLA. Journal of Hydrology 673, doi 10.1016/j.jhydrol.2026.135428.
- Antoniadou N, Stockmarr A, Pedersen JW, Schmith T, Mikkelsen PS (2025): Spatiotemporal analyses of extreme rainfall frequencies in Denmark via Bayesian hierarchical modelling using SPDE with INLA. Stochastic Environmental Research and Risk Assessment 39(9):4125-43, doi 10.1007/s00477-025-03054-5.
The project is one of six DTU Compute projects awarded funding in the latest round of the Villum Synergy programme.
The programme supports projects that combine data science and AI with other fields. Read the press release
The other five projects are:
- PSYFIND: Psychological Fingerprints for Cyber-Deception with Typological Attacker Classification & Adaptive Defense. Associate Professor Emmanouil Vasilomanolakis, DTU Compute – and Professor Ingo Zettler, Department of Psychology – University of Copenhagen.
- Physics-Driven Digital Twins for Two-Phase Cooling in AI Data Centers (SmartCool). Professor Henrik Madsen, DTU Compute and Professor Jesper Hattel, DTU Construct.
- Writing Representation and Probabilistic Trajectory Estimation: Modelling Students’ Writing Development. Associate Professor Michael Riis Andersen, DTU Compute, and Associate Professor Kristine Kabel, The Danish School of Education (DPU) – Aarhus University.
- A Fleet‑Scale Data-driven Framework for Global Wave Sensing (Fleet-GLOWS). Associate Professor Allan Engsig-Karup, DTU Compute and Associate Professor Ulrik Dam Nielsen, DTU Construct.
- Data-driven design of tunable chalcogenide thin films for next-generation indoor and tandem photovoltaics. Associate Professor Mikkel Schmidt, DTU Compute and Senior Researcher Stela Canulescu, DTU Electro.

