Ruf et al. (2026) A fluvial flood risk model for quantifying the benefit of mitigation measures under uncertainty

Identification

• Journal: Natural hazards and earth system sciences
• Year: 2026
• Date: 2026-09-23
• Authors: Mara Ruf, Amelie Hoffmann, Dániel Straub
• DOI: 10.5194/nhess-26-4549-2026

Research Groups

• Engineering Risk Analysis Group, Technical University of Munich, Munich, Germany

Short Summary

This paper presents a dynamic probabilistic flood risk model designed for large-scale, uncertainty-aware flood risk assessments to support decision-making on mitigation measures. The model efficiently couples flood process components and distinguishes between aleatory and epistemic uncertainties, demonstrating its ability to quantify the risk-reduction potential of mitigation measures on the Bavarian Danube.

Objective

• To develop a dynamic probabilistic flood risk model that addresses key challenges in integrated flood risk management, including the need for holistic, large-scale risk assessments, a system-based perspective, and a decision-making framework based on benefit-cost analysis under uncertainty.
• To demonstrate the model's ability to explicitly represent and dynamically couple flood process components, consider aleatory and epistemic uncertainties, and estimate the flood risk-reduction potential of mitigation measures.

Study Configuration

• Spatial Scale: Bavarian Danube, Germany (approximately 380 km river reach). The river and its foreland are laterally discretized into 2 x 892 segments, each 300 to 600 m in length.
• Temporal Scale: Climate-based simulations cover a time horizon from 1980 to 2099, with 3500 years of precipitation data extracted for 1980 to 2050. The flood risk model operates with a temporal resolution of 1 hour.

Methodology and Data

• Models used:
  • Flood Risk Model: Custom-developed dynamic probabilistic model comprising five modules: hydrological load, dike failure, hydrodynamic, inundation, and damage. It is integrated into a 2-level Monte Carlo framework.
  • Hydrological Load: ClimEx project (quasi-random climate-based simulations), WaSim (runoff generation), Larsim (catchment flood routing), and SOBEK (1D non-stationary hydraulic modeling for river channel).
  • Dike Failure: Limit state functions from Vorogushyn (2008) evaluated using Monte Carlo simulation to generate fragility functions.
  • Inundation: ArcGIS for static derivation of inundation elevation-volume (V-E) relationships from a 5 m x 5 m Digital Elevation Model (DEM). Outflow discharge calculated using a modified broad-crested weir equation.
  • Hydrodynamic: Custom vector-based flood routing approach (translational and attenuation vectors) calibrated with 1D hydraulic modeling outcomes.
  • Damage: Basic European Assets Map (BEAM) for land-use classification and asset values, combined with BEAM-specific damage functions for southern Germany.
• Data sources:
  • Climate data from the ClimEx project (IPCC emission scenario RCP8.5).
  • 5 m x 5 m Digital Elevation Model (DEM).
  • Basic European Assets Map (BEAM) for land-use and asset value information.
  • Probability distributions for dike properties based on Vorogushyn (2008) where site-specific data were unavailable.
  • Discharge inputs for flood scenarios provided by Bayerisches Landesamt für Umwelt (LfU).

Main Results

• A computationally efficient dynamic probabilistic flood risk model was developed, enabling large-scale, uncertainty-aware flood risk assessments without runtime coupling of computationally expensive hydraulic simulations.
• The model explicitly represents and dynamically couples flood process components, including downstream flood wave propagation and possible dike failures, through an efficient vector-based flood routing scheme and pre-processed lookup tables.
• A 2-level Monte Carlo framework effectively separates aleatory and epistemic uncertainties, allowing for comprehensive uncertainty and sensitivity analysis at river scales.
• Application to the Bavarian Danube successfully demonstrated the model's ability to estimate the flood risk-reduction potential of a controlled detention basin.
• Sensitivity analysis revealed that uncertainties in the damage module (e.g., damage functions for river sections, business interruption, and indirect damage factors) have the greatest impact on the annual flood risk.
• The uncertainty in the estimated benefit (risk reduction) of mitigation measures is significantly smaller than the uncertainty in the overall flood risk, due to high correlation of uncertainties with and without the measure (coefficient of variation of risk reduction is 25%).
• Scenario analysis indicated that an increased flood occurrence rate due to climate change leads to a larger expected flood risk reduction from mitigation measures but also results in increased uncertainty.

Contributions

• Introduces a novel, computationally efficient probabilistic flood risk modeling framework that effectively balances process representation, spatial extent, and comprehensive uncertainty analysis.
• Provides a unique approach for dynamically coupling flood process components (e.g., dike failures, mitigation measures, and downstream impacts) without requiring computationally intensive runtime simulations.
• Enables a robust, uncertainty-aware evaluation of the benefits of flood mitigation measures through a two-level Monte Carlo framework that explicitly separates and quantifies aleatory and epistemic uncertainties.
• Offers a practical tool for large-scale probabilistic risk assessments and supports decision-making in integrated flood risk management, including benefit-cost analyses and the joint optimization of mitigation portfolios at river scale.
• Highlights the critical importance of damage module uncertainties in overall flood risk estimates, guiding future research and development priorities.

Funding

• Bayerisches Landesamt für Umwelt (LfU)
• Bayerisches Staatsministerium für Umwelt und Verbraucherschutz (StMUV)

Citation

@article{Ruf2026fluvial,
  author = {Ruf, Mara and Hoffmann, Amelie and Straub, Dániel},
  title = {A fluvial flood risk model for quantifying the benefit of mitigation measures under uncertainty},
  journal = {Natural hazards and earth system sciences},
  year = {2026},
  doi = {10.5194/nhess-26-4549-2026},
  url = {https://doi.org/10.5194/nhess-26-4549-2026}
}