Author(s): Min Xu; Kai Wang; Shunfeng Peng; Karel Heynert; Eric Sprokkereef; Gerard De Vries
Linked Author(s): Min Xu
Keywords: Delft-FEWS; Flood forecasting; Huaihe River basin; Open source
Abstract: In order to minimize the damage during a flood event, on the one hand, it needs a proper decision which relies on accurate flood forecasting. On the other hand, the lead time is important for people to take actions. However, flood forecasting usually involves a complex process which is time consuming, especially when multiple data sources and models are used. This paper introduces open source software, Delft-FEWS, developed by Deltares in the Netherlands. Delft-FEWS is a platform to integrate multiple forecasting models into one operational system as well as for data management. It is fully configurable based on the requirement of each application. This gives users complete freedom to adjust their systems or to implement more functions. Operational forecasting can also be automated in Delft-FEWS which facilitates the forecasting procedure and reduces the forecasting time. Different results can be disseminated in Delft-FEWS, for example time series in figures and spatial data in animations. In the test case, Delft-FEWS is applied to the Huaihe River basin in China. It connects the online database and preprocesses the data. Several models are implemented in the basin using these data, for example the Xinanjiang model and the HEC-RAS model for both hydrological and hydraulic forecasting, respectively. The application demonstrates that Delft-FEWS is a useful and efficient tool for flood forecasting. In the future research, hydraulic modeling of the low Huaihe River is required due to the flatness and the hydrological modeling needs to cover the whole basin. More important, they are going to be integrated into Delft-FEWS and probabilistic forecasting is needed to deal with uncertainty.
Year: 2013