Surface water and urban flood modelling

What we can help you with

Intense rainfall can cause water to accumulate in depressions and find new routes through built-up areas. Terrain, buildings and other obstacles influence where water flows, while infiltration and drainage affect how much remains on the surface. Computational tools help investigate these relationships and assess the consequences of different measures.

We work with municipalities, consultants, developers and technology providers to:

• Identify catchments, surface flow paths and areas where water can accumulate.
• Analyse how rainfall, infiltration and changes to terrain and buildings affect surface water.
• Develop fast tools for comparing scenarios and mitigation measures.
• Develop and adapt dynamic models of flow and inundation.
• Integrate computational methods into planning and decision-support software.

Contact our Applied Computational Science research group to discuss an analysis task, a development project or the need for new models and tools.

Fast terrain analysis with SWIM

SWIM – Surface Water Integrated Modeling is our open-source software for terrain-based surface water analysis. It identifies catchments, drainage paths and depressions, and describes how areas of accumulated water connect as water levels rise and depressions overflow.

The methods are computationally inexpensive and support interactive exploration of large terrain areas. They make it possible to compare alternatives and investigate how buildings, obstacles, drainage and terrain modifications affect flow paths.

SWIM also includes simplified infiltration and calculations of how water accumulates and drains over time during rainfall events. These calculations use terrain analysis and water balances. When flow velocities and rapid changes in water level are important, dynamic flow models are needed.

Dynamic simulation and interaction between models

We develop numerical methods for the shallow-water equations, which describe how water depth and flow evolve over time. Our work includes efficient computation on graphics processing units (GPUs), wetting and drying, and adaptive mesh refinement where greater detail is required.

We are redeveloping our earlier dam-break and flood simulator in Julia and working towards closer interaction between terrain analysis and dynamic simulation. Fast analyses can provide an overview and identify areas requiring further investigation, while dynamic models can examine the flood event itself. For coastal areas, our experience also includes storm surge modelling through GPU Ocean.

From early GPU research to practical planning tools

Our work builds on more than twenty years of research into numerical methods for flow. In 2003, we began investigating how graphics cards could be used to solve partial differential equations. Shallow-water equations and dam-break problems provided early examples of combining fast computation with visualisation.

This developed into a collaboration with the National Center for Computational Hydroscience and Engineering (NCCHE) at the University of Mississippi on GPU-accelerated simulation of dam-break flooding. The software was subsequently adapted for urban flooding and licensed to a UK engineering consultancy. The collaboration brought together method development, verification and validation, and transfer into practical use.

A second research line began with algorithms for identifying structural traps for underground CO₂ storage. These use a direct analogy with watershed analysis: water collects in depressions and spills through outlets, whereas CO₂ moves upwards and becomes trapped beneath elevated parts of an impermeable caprock.

Building on these ideas and results from Voldsund's 2017 master's thesis, we adapted the algorithms with Spacemaker for urban surface water analysis. The methods were integrated into the company's planning software through the OptiSite collaboration. This research line continues in SWIM and demonstrates how mathematical ideas can be transferred between application areas.

Open software and further development

SWIM is written in Julia and is available with open source code, documentation and examples. The software provides a foundation for further development and integration into other tools.

Development continues through activities including SUrbArea, which develops and tests surface water solutions for sustainable urban development, and builds on work undertaken in SWAMP.

Publications and further reading

• Andersen (2025). Topography-based surface water modeling in Julia, with support for infiltration and temporal developments. Journal of Open Source Software.
• Sætra, Brodtkorb and Lie (2015). Efficient GPU-implementation of adaptive mesh refinement for the shallow-water equations. Journal of Scientific Computing.
• Sætra (2014). Shallow water simulation on graphics hardware. PhD thesis, University of Oslo.
• Brodtkorb, Sætra and Altinakar (2012). Efficient shallow water simulations on GPUs: Implementation, visualization, verification, and validation. Computers & Fluids.
• Hagen et al. (2007). How to Solve Systems of Conservation Laws Numerically Using the Graphics Processor as a High-Performance Computational Engine.
• Gemini: Applying oil and gas expertise to urban flood prevention (Norwegian).
• Tekna: Artificial intelligence and surface water challenges (Norwegian).
• Early demonstration of real-time simulation of dam breaks, tsunamis and storm surges.