Spillway design floods In Sweden: II. Applications and sensitivity analysis

Open Access
Publisher Website
Google Scholar
Abstract
The recently developed guidelines for design flood calculation in Sweden are based on simulation with a hydrological model. Several meteorological and hydrological conditions are prescribed as input to the model. The sensitivity of the floods according to the new guidelines with respect to those prescriptions is checked as well as the sensitivity to the modelling procedure. In this paper, the HBV hydrological model is used. It is shown that the most sensitive input factor is the design 14-day precipitation sequence. For autumn floods, the scaling relation between the precipitation sequence and the generated floods is in the order of 1:1. For spring floods on the other hand, snowmelt influences the flood magnitude and reduces the sensitivity to precipitation to about 1:0.6. Soil moisture modelling was important in basins with high evapotranspiration. Model calibration could have a large effect on the design flood magnitude. The most sensitive parameter was the high flow recession coefficient, K0 . The water stage development in a multi-reservoir system is further influenced by the regulation strategy and spillway capacity. An increased spillway capacity can improve the situation considerably for a particular dam but could at the same time worsen the situation for downstream dams. After the filling of the reservoir there was a clear relation between inflow peak and maximum water stage. The study shows that it is difficult to assess the integrated effects of extreme precipitation, snowmelt, soil moisture status and regulation in a system beforehand.