Performance evaluation and improvement of river flood simulation models for the Demer basin, Belgium
Bokkor Siddik, A. (2010). Performance evaluation and improvement of river flood simulation models for the Demer basin, Belgium. MSc Thesis. Katholieke Universiteit Leuven (KUL)/Vrije Universiteit Brussel (VUB): Brussel. X, 81 pp. |
| Available in | Authors | | Document type: Dissertation
|
| Keywords | Evaluation
Floods
Hydrology models
Modelling
Performance evaluation
Physics > Mechanics > Fluid mechanics > Hydrodynamics
Sensitivity
Simulation models
Uncertainty
Water bodies > Inland waters > Rivers
Belgium, Demer Basin [Marine Regions]
Fresh water |
| Authors | | Top | - Bokkor Siddik, A.
- Van Steenbergen, N., revisor
| | |
| Abstract | Floods are the most costly natural disasters in terms of human sufferings and economic losses all over the world. Although flood simulation models are essential to minimize flood losses, evaluation of this kind of models for assessing different uncertainty sources is still a major concern.A systematic evaluation of flood simulation model provides insight into the level of confidence in model estimates and can aid in assessing how various possible model estimates should be weighed. It yields the chance of evaluating the impact of model improvement actions. In addition, it can lead to the identification of the key sources of uncertainty which merit investigation for further improvement. Hence before their use as decision making aid, models must be properly evaluated to improve their prediction accuracy and to reduce the likelihood of making decisions that could lead to undesirable outcome. In this study, model performance evaluation has been carried out for two flood simulation models of the Demer basin, Belgium to investigate accuracy of these models. To analyze the rainfall uncertainty, evaluation has been performed by comparing original (existing) and optimized rainfall inputs, simulating these inputs in rainfall-runoff models and, further on, the rainfall-runoff results in river hydrodynamic models. The performance of the rainfall-runoff model and hydrodynamic results has been evaluated by means of statistical indices and graphical interpretation. To achieve this task, two types of lumped conceptual rainfall-runoff models and two types of river hydrodynamic models have been considered for the study: PDM linked to InfoWorks-RS software, and NAM, linked to MIKE11 river modelling software. The differences in the results gave information about the importance of the rainfall input accuracy, the choice and the calibration results of the rainfall-runoff model and the choice of the hydrodynamic model. Also the effects of some model improvement actions for the hydrodynamic model were investigated. This study deals with only off-line flood simulation models (without real time flood forecasting). |
|
