Sabeti et al. (2026) A Monte Carlo-based approach for reconstructing past flood events

Identification

• Journal: Hydrological Sciences Journal
• Year: 2026
• Date: 2026-07-07
• Authors: Ramtin Sabeti, Thomas Kjeldsen, Ioanna Stamataki
• DOI: 10.1080/02626667.2026.2685848

Research Groups

• Department of Architecture and Civil Engineering, University of Bath, UK
• JBA Consulting, Skipton, UK
• School of Engineering, University of Greenwich, Chatham Maritime, Chatham, UK

Short Summary

This study presents a new Monte Carlo-based approach for reconstructing past flood events using 2D hydraulic modelling. The method is applied to the 1968 River Chew flood in southwest England and provides an empirical probability distribution of peak flow.

Objective

• Reconstruct historical flood events with associated uncertainty analysis
• Develop a Monte Carlo-based inverse modelling approach for 2D hydraulic modelling

Study Configuration

• Spatial Scale: Local scale (River Chew catchment, UK)
• Temporal Scale: Past event reconstruction (1968)

Methodology and Data

• Models used: HEC-RAS 2D model
• Data sources:
  • Historical flood marks on buildings
  • Gauging station data (1958-2020)
  • Satellite imagery and aerial photographs

Main Results

• The Monte Carlo approach generated 1000 input parameters sets, varying Manning’s n values and hydrographs.
• A subset of simulations matched the observed flood marks within an acceptable range of ±20 cm.
• The peak flow corresponding to these experiments was summarized in a histogram, showing that the most frequent peak flow was between 156 and 175 m3/s.

Contributions

• This study provides a novel Monte Carlo-based approach for reconstructing past flood events using 2D hydraulic modelling.
• The method integrates uncertainty analysis with 2D hydraulic modelling, allowing for more accurate reconstruction of historical flood events.

Funding

• Leverhulme Trust [Project Grant: RPG-2022-306]
• Google Cloud Research Credits [GCP19980904]

Citation

@article{Sabeti2026Monte,
  author = {Sabeti, Ramtin and Kjeldsen, Thomas and Stamataki, Ioanna},
  title = {A Monte Carlo-based approach for reconstructing past flood events},
  journal = {Hydrological Sciences Journal},
  year = {2026},
  doi = {10.1080/02626667.2026.2685848},
  url = {https://doi.org/10.1080/02626667.2026.2685848}
}