Internship Colloquium Caspar Cluitmans

In this research the brook “Voer” was modelled with a combination of the hydrodynamic model SOBEK and the rainfall-runoff model HBV. The model performance was analysed and compared with an older model.

Organised by Hydrology and Quantitative Water Management

Thu 5 February 2015 13:30 to 14:00

Venue Lumen, gebouwnummer 100
Room Lumen 2

Modelling the brook “Voer” and analysing model performance.

In the past water problems occurred in the “Voer” catchment. These were partly caused by a fast reacting catchment and a canalised brook. The goal of the research was to create an up-to-date model for the brook “Voer”, to analyse the performance of this new model and to compare it with an older model (which consisted of the hydrodynamic model InfoWorks RS and the rainfall-runoff model PDM). Two models were created for this goal: 1) an up-to-date model called “SOBEK-B” consisting of HBV and SOBEK-Rural. HBV was used to model the rainfall-runoff process in the catchment and its output was used as input for the hydrodynamic SOBEK-Rural model, which was used to model the brook itself. 2) A hydrodynamic SOBEK-Rural model called “SOBEK-A” which was used to model the brook and was structured in the same way as the old model, so that the performance of the two models could be compared. It was difficult to calibrate HBV for the “Voer” catchment. This was probably caused by the complex geology of the catchment in combination with the simplified representation of the slow-flow reservoir in HBV. Literature research showed that there are empirical conceptual models that can simulate rainfall-runoff processes well in catchments with similar complex geology. With the obtained optimal parameter-set HBV did not simulate the catchment well. For the calibration year a NS of 0.64 was obtained and for validation a NS of 0.11. During the manual calibration correlation was detected between the fc and rfcf parameters. The “SOBEK-B” model contained three retention basins, in contrary to the “SOBEK-A” model and the old model. The “SOBEK-B” model did not simulate the brook that well, partly due to the mediocre quality of the input originating from the HBV simulation. The addition of the hydraulic component (SOBEK) to the hydrological component (HBV) clearly led to dampening of the discharge peaks. This dampening effect was mainly caused by the retention basins. A sensitivity analysis showed that the kmanning parameter was sensitive. The “SOBEK-A” model used exactly the same input as the old model. The model simulated the discharges lower and the water heights higher than the old model. It was difficult to assess if the “SOBEK-A” model performed better than the old model due to the differences in the used model structures.