A physically-based model approach to assess the effectiveness of single and combined measures of decentralized flood protection

作者：Rieger Wolfgang^*; Disse Markus

来源：Hydrologie und Wasserbewirtschaftung, 2013, 57(1): 14-25.

DOI：10.5675/HyWa_2013,1_2

• Decentralized flood protection measures
• land use change
• process based simulation
• restoration of rivers
• small retention ponds
• soil measurement
• soil tillage

This paper deals with process-based simulations of distributed flood-control measures such as changes of land-use or soil-tillage practices, small retention ponds, and the rehabilitation of rivers aiming to quantify the flood-reducing effects. In this context several challenges have to be met, because these measures have different impacts on the whole rainfall-runoff process including runoff generation, runoff concentration, and flood routing. This study chooses a combined modelling approach that comprises the physically based rainfall-runoff model WaSiM-ETH (Version 8.4.2, SCHULLA 1997) and the 2D hydrodynamic numerical model HYDRO_AS-2D (NUJIC 1998) that is parameterized also by means of measurements in the field. On the one hand, these field data prove the influences of the different types of land uses (grassland, forest, ploughland) on soil hydraulic properties and provide the basis for the land-use-dependent parameterization of the soil model in WaSiM-ETH. The measurements show that the soil matrix under grassland has the best infiltration and water storage properties among the types of land use considered. On the other hand, in combination with laboratory studies, the collected data are used to identify the best fitting pedotransfer function for the study area, which is importantto minimize uncertainties regarding the use of different types of pedotransfer functions in a physically based rainfall-runoff-model. After being constructed, parameterized, and calibrated, the models are coupled with an offline-connection by adding the runoff of each sub-basin in WaSiM-ETH to HYDRO_AS-2D as boundary conditions. The high fitting quality of calibration results demonstrates that the chosen coupled modelling approach can be applied in further research. The distributed flood control measures are parameterized by using field and laboratory data, literature sources, and historical marshland and drainage maps. Subsequently their flood-reducing effectiveness is quantified in the rural, mesoscale study area of the Windach catchment (A(EZG) 65 km(2)) for different flood events. The simulation results show that the effectiveness of flood-reducing measures is influenced on a case-by-case basis by different parameters, e.g. the course of the hydrograph, the peak flow or the runoff volume. The highest effect is achieved by an integrated concept including retention ponds, rehabilitation of rivers and marshlands as well as afforestation. In this case, peak flow can be reduced by 11 % (HQ(60) advective event) to 26% (HQ(60) convective event).The small retention ponds (max. 50,000 m(3)) turned out to be the most effective single measures of distributed flood control as they can reduce the flood peak even of an advective HQ(100) by roughly 10%. Generally, decentralized flood control has the highest potential in the case of convective heavy rainfall events.

• 出版日期2013-2