University of Twente Student Theses
Evaluation of ensemble flood forecast performance by using a semi-distributed multi-model approach compared to single models
Groot, S.A. de (2025) Evaluation of ensemble flood forecast performance by using a semi-distributed multi-model approach compared to single models.
|
PDF
60MB |
| Abstract: | The increasing frequency of extreme precipitation events presents significant challenges to flood preparedness and mitigation globally. Reliable and accurate Flood Forecast and Warning Systems (FFWS) are crucial for ensuring flood safety. A key component of a FFWS is a hydrological model that translates precipitation into runoff. Instead of relying on a single hydrological model, integrating multiple model structures offers a promising approach to enhancing FFWS performance by capturing a broader range of uncertainties. This study evaluates and compares the flood forecasting performance of two semi-distributed hydrological models HBV and GR4H versus a model combination consisting of both, for the Vecht River in Germany and the Netherlands. Both models showed good forecast performance for the upstream (independent) sub-catchments. Forecast accuracy for downstream (dependent) sub-catchments was poor at short lead times (<12 hours) for both HBV and GR4H. This was caused by inaccurate state updating, discrepancies between observed and simulated discharges, and uncertainty in observed discharge from catchments upstream. The multi-model forecasts showed marginal improvements based on evaluation metrics, with respect to the individual model forecasts. In addition, the multi-model forecasts substantially outperformed the weaker individual model, therefore reducing the possibility of a poor forecast. |
| Item Type: | Essay (Master) |
| Clients: | Deltares, Delft, Netherlands |
| Faculty: | ET: Engineering Technology |
| Subject: | 43 environmental science, 56 civil engineering |
| Programme: | Civil Engineering and Management MSc (60026) |
| Link to this item: | https://purl.utwente.nl/essays/104945 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page