The role of hydropower is expected to change in the close future in the Nordic countries. In order to fulfill targets set by the Paris agreement, the reliance on low emission intermittent power production such as solar and wind power is likely to increase. This in turn means that the operating conditions of Nordic hydropower plants are expected to change to balance the volatility of intermittent power production. Hence, rapid flow fluctuations due to changes in power demand, so called hydropeaking, is likely to increase in the future. Hydropeaking is associated with many negative impacts on the reach downstream of the hydropower plant. It is with this background that the research consortium HydroFlex was created. One of the aims of HydroFlex is to investigate the effects of very high frequency hydropeaking on adjacent river reaches. In this work, several model scenarios involving high frequency hydropeaking in a bypass reach in the river Umeälven were investigated. The river dynamics was modeled using the solver Delft3D. By using LIDAR measurements in the winter, when the reach was mostly dry, it was possible to capture the bathymetry with high spatial resolution. A 2D model was then set up using the measured bathymetry and calibrated using eight pressure sensors placed along the reach. The transient response in the reach was then compared between the measured depth from the sensors and the model (Paper A). Once the calibration had been performed, the effects of an increase in hydropeaking frequency on the downstream reach, could be investigated. A hysteresis effect in the depth increase-decrease cycle was observed in both measurements and simulations. Additionally, high hydropeaking frequency led to a hydraulic stage in the river where steady state is never achieved (Paper B). This effect, and how it might affect downstream salmonid spawning areas, was then further investigated. It was observed that an increase in hydropeaking frequency could reduce the proportion of potential spawning areas in the downstream reach that would dry out. Furthermore, a general method of computing dewatering time for numerical models was provided. This method was then applied, in tandem with a stranding model, to investigate how different spill gate closing times affected salmonids propensity to stranding. The likelihood of stranding varied between species and life-stage. Moreover, the tendency of stranding decreased longitudinally downstream (Paper C). The findings of a hydraulic stage in (Paper B) merited further investigation. By investigating model scenarios, where the hydropeaking frequency varied between 10 and 80 flow changes per day, four different hydraulic stages that occur in the reach, were identified. A method of classifying the hydraulic stages was suggested. The identified hydraulic stages were: the dewatering stage, the alternating stage, the previously identified dynamic stage and lastly the uniform stage. The relation between the hydraulic stages and the hydropeaking frequency were then investigated using Fourier analysis. It was observed that the hydropeaking frequency is the dominant variable in the dynamic and uniform stages, while the alternating stage is more complex. Additionally, the dynamics of the stages could be predicted by performing a Fourier transform of one increase-decrease cycle. The Fourier spectrum could also be used as an alternative way of classifying the stage (Paper D). Numerical methods on their own is not a holistic tool to predict river flows, they require field measurements for calibration and validation. In (Paper E) an overview of the methods used in the field in this work is given. Moreover, a discussion on the limitations of 2D modeling for large scale river applications is presented.