Dessers et al. (2026) Hydrological modelling of the 2021 mega-flood in the east of Belgium

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-01-02
• Authors: C. Dessers, P. Archambeau, S. Erpicum, B. Dewals, M. Pirotton
• DOI: 10.1016/j.jhydrol.2025.134901

Research Groups

• Research Group of Hydraulics in Environmental and Civil Engineering (HECE), University of Liège, Belgium.

Short Summary

This study reconstructs the 2021 mega-flood in Belgium's Vesdre and Amblève catchments using the WOLFHydro platform, demonstrating that models calibrated on historical data significantly underestimate extreme peaks. The research highlights the necessity of flexible modeling frameworks and envelope curves to account for the high uncertainty in "black swan" hydrological events.

Objective

• To perform an event-based hydrological analysis of the July 2021 floods in the most impacted Belgian catchments to reconstruct missing hydrographs and evaluate the reliability of different modeling approaches (gridded vs. lumped) for extreme events.

Study Configuration

• Spatial Scale: Catchment scale (Vesdre and Amblève catchments, East Belgium).
• Temporal Scale: Event-based analysis (July 2021) with comparisons to historical flood data.

Methodology and Data

• Models used: WOLFHydro (in-house computationally efficient gridded runoff model), GR4H (lumped conceptual model), VHM (lumped conceptual model), and 2D hydraulic simulations for flood extent validation.
• Data sources: Quantitative Precipitation Estimates (QPE), gauging station records (where available), and field observations of flood extents.

Main Results

• Reconstructed missing hydrographs for the Vesdre catchment, which received approximately 175 mm of rainfall, using a gridded runoff model coupled with 2D hydraulic simulations that matched field observations.
• In the Amblève catchment (120 mm of rainfall), models achieved high accuracy (NSE = 0.89–0.99) when calibrated specifically for the 2021 event.
• Models calibrated on historical floods underestimated the 2021 peak discharges by up to 45%.
• Conversely, models calibrated on the 2021 extreme event performed inconsistently when applied to less extreme historical events.
• Rainfall input accuracy was identified as the most critical factor influencing the quality of the simulated hydrographs.

Contributions

• Development and application of a new, computationally efficient gridded runoff model within the WOLFHydro platform.
• Reconstruction of critical missing hydrological data for the Vesdre valley, where gauging stations failed during the 2021 event.
• Evidence-based critique of traditional calibration strategies, showing they are often inadequate for predicting extreme "mega-floods."
• Provision of a flexible modeling framework (WOLFHydro) capable of integrating different model structures for robust flood analysis.

Funding

• Not specified in the provided text.

Citation

@article{Dessers2026Hydrological,
  author = {Dessers, C. and Archambeau, P. and Erpicum, S. and Dewals, B. and Pirotton, M.},
  title = {Hydrological modelling of the 2021 mega-flood in the east of Belgium},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2025.134901},
  url = {https://doi.org/10.1016/j.jhydrol.2025.134901}
}

Generated by BiblioAssistant using gemini-3-flash-preview (Google API)