Graduation of Roline Montijn

Numerical study on the development of alternate bars in varying discharge conditions

• Professor of graduation: Dr.ir. Astrid Blom

• Supervisors of graduation: Dr. ir. E. Mosselman (Deltares & TU Delft), dr. ir. C.J. Sloff (Deltares & TU Delft), dr. ir. R.J. Labeur (TU Delft), dr. R. Schielen (TU Delft)

River bars are large-scale bedforms formed by the local deposition of sediments, the length of which scales with the channel width and the height with the water depth.  They create suitable habitats for aquatic fauna and riparian vegetation,  which makes them useful in river restoration projects.   Understanding the dynamics of river bar development is required for a proper design of the restoration project and for developing a sustainable management scheme. The  channel  width-to-depth  ratio  is  the  key  controlling  parameter  for  the  formation  of  river  bars. Existing stability analyses define a resonance point and critical width-to-depth ratio, which mark the transition between different types of bar regimes. By increasing the width of a river section to a width-to-depth ratio above its critical value, free, migrating bars are expected to form.  When they are fixed to a local perturbation, the bars show a periodic pattern, with an amplitude damped in downstream direction.  By further increasing the width-to-depth ratio towards its resonance point, the celerity of the free bars goes to zero. This results in a pattern of steady bars. Due to the natural variability of river discharge, the width-to-depth ratio varies over time.  The objective of this research is to give insights into the development of river bars in varying discharge conditions. For that, insight in the timescale of adaptation of the river bars to new flow conditions is necessary .A straight river channel is modelled in Delft3D, with non-erodible banks to which a fixed perturbation by means of a groyne is added.  The geometry of the river model and the discharge hydrograph are roughly based on the Dutch river Grensmaas.  In the first set of simulations, the upstream boundary condition is a constant discharge, with magnitudes between 50 and 1250 m3/s.  From this constant-discharge analysis,  the different types of river bars and the time to develop the bars are analyzed. The transition between the river bar regimes, around the resonance point, is modelled by means of a steadily increasing and decreasing discharge, starting with a fully developed bed.  The response of the river bars and the transition of the river bar regimes due to the varying discharge are assessed and coupled to the timescale of development based on the constant-discharge analysis and the timescales of flow unsteadiness due to natural discharge variability. From the constant-discharge analysis,  three river bar regimes are determined,  being superresonant, subresonant  and  stable,  ordered  in  increasing  discharge.   Focussing  on  the  bars  fixed  to  the  local perturbation,  the  hybrid  bars,  we  see  a  pattern  of  bars  with  an  amplitude  growing  in  downstream direction, a pattern of bars with an amplitude damped in downstream direction or the absence of bars. The damping length and wavelength of the bars show good agreement with the values determined from the theory of Struiksma et al. (1985).  Both in superresonant as in subresonant conditions, free bars developed in the domain;  in superresonant conditions, they became suppressed as a pattern of hybrid bars developed. During subresonant conditions, free bars developed at the end of the pattern of bars, when the amplitude of the pattern of hybrid bars goes to a few centimetres. No free bars develop in the stable bar regime. The transitions between the three regimes are marked by the resonance point and the critical width-to-depth ratio.  The resonance point is between superresonant conditions and subresonant conditions, and is clearly obtained from the numerical model by analyzing the damping lengths of the pattern of hybrid bars.  It is assigned to a width-to-depth ratio of 37-38.  Contrary to existing stability analyses, the critical width-to-depth ratio for the formation of free bars is marked by a gradual transition instead of a single value. In this transition, free bars develop with intervals of a few months. This transition is at width-to-depth ratios from 11 to 20. Three timescales have been defined with respect to the development of bars.  First, the development of a group of free bars gives the timescale of free bars.  This timescale shows little variation between the different simulations, and stays around 20 days.  Secondly, the time to develop the first bar in the pattern of hybrid bars is up to 10 times greater than the timescale of free bars.  Thirdly, the timescale to reach the final bed topography is in the superresonant and subresonant bar regime the time towards a pattern of hybrid bars.  Based on the definition of the damping length by Struiksma et al. (1985), a theoretical expression is defined, which gives a reasonable fit to the third timescale. The distinction between timescales of hybrid and free bars results in a different response of free and hybrid bars to a varying discharge.  The first response of the bed is the adaptation of the free bars towards the new bar regime.  A pattern of hybrid bars follows, developing in downstream direction from the fixed perturbation.  A type of hysteresis in the transition around the resonance point of the river is detected, due to an increased stability of the bed when a hybrid bar pattern has developed. This bed stability suppresses the formation of free bars in the transition from superresonant to subresonant conditions. This thesis assesses the timescales of river bars and the transitions between the river bar regimes. The longer timescales for lower discharges suggest that subresonant conditions are more determinative for the final bed topography than superresonant conditions, which corresponds to the development of migrating bars. But this thesis also shows the increased stability of the bed due to the formation of alternate bars in superresonant conditions.  Therefore, the presence of migrating bars is not solely related to discharge conditions in subresonant conditions, but the bed topography from earlier discharge conditions should be taken into account. This study also shows the suitability of numerical models to study short- and long-term processes, as they clarify laboratory, theoretical and field processes.