Assessment of the structural uncertainty of hydrological models and its impact on flood inundation mapping

ABSTRACT

Floods are frequent natural disasters, and the management of floods involves predicting flood hydrographs and inundation maps. However, the uncertainty associated with these predictions is a significant challenge in flood management. This study focuses on the uncertainty associated with the selection of mathematical models to estimate flood hydrographs and their impact on inundation map predictions. The study involved setting up 27 hydrological models using different mathematical models for the loss, transform, and routing methods within the Hydrologic Engineering center – Hydrologic Modelling system (HEC-HMS) model. The results showed that the peak flow estimation depends on the models adopted for predicting the flood hydrograph. This uncertainty was translated to the flood inundation maps, where the inundated area varied across different flood events. The study also found that the hydrological conditions prior to the start of the event impact the hydrological model predictive uncertainty, and ensemble predictions by Bayesian model averaging can reduce the uncertainty.

KEYWORDS:

• hydrological modelling
• structural uncertainty
• Bayesian model averaging
• containing ratio
• deviation in amplitude
• flood inundation

Editor A. Fiori; Associate Editor S. Persiano

Editor A. Fiori; Associate Editor S. Persiano

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.