Research projects
Project background
The flood protection project Alpine Rhine International reach (external page www.rhesi.org) aims at increasing the discharge capacity between km 65 (confluence with the tributary Ill) and km 91 (mouth into Lake Constance) from currently 3'100 m3/s (approx. 100-year flood) to 4'300 m3/s (approx. 300-year flood) within the next decades.
This increase in discharge capacity is to be achieved without raising flood levels or heightening the flood protection dams. One of the main measures of the project thus consists in increasing the discharge cross-section by widening the middle channel and extending the flow dynamics into parts of the floodplains. The characteristic double trapezoidal profile of the Alpine Rhine, which is characteristic for the international stretch and was established in earlier corrections between 1892 and 1923, will be abandoned over long sections of the river (Fig. 1). These adaptations of the channel geometry are intended not only to increase flood safety but also to enable a near-natural design of the channel.
VAW tasks
The planned river engineering measures will be verified and optimized by VAW on behalf of the external page International Rhine Regulation (IRR). The following research topics are being investigated by hybrid modeling, i.e. parallel application of physical and numerical modeling:
- Flood safety
- Morphological development
- Processes caused by changes of the river cross-sections (narrowing / widening)
- Bank protection and embankment safety
- Bridge structures (freeboard, risk of log jams, pier scours)
- Vegetation on gravel banks (backwater effects during floods and retention potential of driftwood, influence on morphological development)
Experimental program
The investigations are carried out in an external building (Abb. 2) in two physical models with a scale of 1:50 for characteristic sections of the project:
- The narrowest section of the project is located in the section "Widnau / Höchst" (km 81.1 to 86.4). This section is suitable for investigating the processes due to cross-section changes (narrowing, widening) as well as stress on river banks in river bends.
- The section "Oberriet / Koblach" (km 66.0 to 71.1) contains the largest widening within the project. The transition from the present narrow riverbed with widths of 70 m to a braided channel with widths of up to 350 m will be carried out by removing the bank protections and allowing a dynamic development of the river reach. The investigation of the morphological dynamics is of great interest with regard to ecology and bedload transport.
In addition to the models in question, hydronumerical 2D bed-load models were run in the BASEMENT simulation software, which represent the entire perimeter of the Alpine Rhine. These are elementary bases for providing and checking the boundary conditions of the models in question, validating the results and carrying out sensitivity analyses (“hybrid modeling”). The numerical simulations were further supplemented with morphodynamic calculations. The numerical parameters were calibrated and validated using the model in question. As a result, the bed changes during various discharge scenarios could also be simulated very well in the numerical model.
Widnau/Höchst section
The investigations of the first section “Widnau / Höchst” were carried out between June 2019 and January 2021. After calibrating the present model in the current river engineering state, investigations were carried out into the development of the morphology after implementation of the project. These investigations were intended to show where the bank protection should be supplemented or optimized with additional protection and guidance structures (Fig. 3). The tests were concluded with investigations of driftwood and the probability of their blockages at bridge piers.
Oberriet/Koblach section
The Oberriet/Koblach section was investigated between January 2021 and the end of 2023, with the model setup taking around 4 months and the model tests lasting 1.5 years.
In this section, the focus was on the transition from the monotonous, narrow existing channel to a dynamic, very wide channel. The widening should take place within a few years with minimal mechanical effort and as dynamically as possible. The model was used to investigate whether this transformation is possible without mechanical effort and which initialization measures need to be implemented.
The first series of tests showed that the removal of bank protection alone is not sufficient to reshape the channel within a few years. In the further course of the project, various measures were tested to initiate self-dynamic development. These were then combined by the planners into an overall concept, which was again tested in the model. In this concept, mechanical interventions only take place during an initial construction phase; thereafter, the channel develops largely on its own. Within a few years, the development of the bed could be observed with this concept (Fig. 4).
The results of both investigations will be shown in the Paper «Wasser Energie Luft, WEL» issue 116 (2) from June 2024.
Wood is an important component of a river ecosystem, as it can create heterogeneous flow conditions and morphological structures. If the amount of wood in a river exceeds the river transport capacity, the risk of wood accumulations can increase at river infrastructures such as bridges or weirs. Wood accumulations at river infrastructures can lead to an increase in water depth upstream of the structure (i.e., backwater rise), which may result in flooding of the surrounding area or to structural damage. Large wood, therefore, may represent an additional hazard potential during floods. To study the hazards related to large wood in rivers, a series of flume experiments was conducted to quantify the accumulation probability of large wood at bridge piers (PhD thesis Download Schalko, 2018 (PDF, 33.9 MB)). The results of the flume experiments demonstrated that the accumulation probability increases with increasing log length and decreasing approach flow velocity. Additional field tests were conducted to evaluate the upscaling of the flume experiments and to further improve the process understanding of the accumulation as well as the wood-structure (bridge pier) interaction.
The field test was conducted in collaboration with the Maintenance section of the Canton Zurich at the River Glatt. The idea was to study the accumulation process of large wood at a circular, single bridge pier under prototype conditions (Fig. 1). The test setup was defined based on the conducted flume experiments at VAW. Individual logs (Length L = 4 m) were added to the River Glatt perpendicular to the flow direction and 5 to 10 m upstream of the bridge pier (d = 1.2 m) using a crane truck. Similar to the flume experiments, this procedure was repeated 40 times to obtain a statistically significant accumulation probability. In addition, the accumulation process was analyzed using camera and video footage.
Wood is a relevant part of a river ecosystem and affects both flow conditions and morphological structures. In literature, wood is commonly referred to as large wood (LW), defined as logs with a diameter ≥ 0.1 m and a length ≥ 1.0 m. Transported LW in rivers may lead to accumulations (logjams) at shallow water areas or at river infrastructures (e.g., bridges or weirs). Such logjams generate important riverine habitat by increasing the upstream water surface elevation, i.e., backwater rise, and creating an upstream pool with slower, deepened water. Depending on the number of transported logs and the flow conditions, resulting backwater rise can provoke a flood hazard leading to inundation or structural damage. Therefore, the prediction of backwater rise due to logjams is required to inform river restoration as well as flood hazard assessment efforts.
Existing approaches either consider collections of individual logs, which cannot describe large logjams, or are based on empirical equations (Schalko et al., 2018) and therefore only valid for the investigated parameter range. Follett et al. (2020) demonstrate that logjams comprised of many logs act as a porous structure that generates momentum loss proportional to the number, size and packing density of the logs and the length of the jam. Backwater rise is predicted from unit discharge, unobstructed water depth, and a dimensionless jam structural parameter. This structural parameter comprises log number, size, and packing density, and can be found from a set of measurements of river discharge and water depth upstream and downstream of the jam. The new logjam model can further be used to describe how the length of the backwater rise, associated pool size, and the increase of sediment deposition depend on jam structure and channel slope. The current model to describe backwater rise due to logjams was derived based on channel-spanning logjams (Figure 1). In nature, logjams can exhibit various shapes, including partially spanning logjams (Figure 2). The associated hydrodynamic processes have not been studied using flume experiments. Therefore, the main questions of this project are:
- How are flow depth and flow velocity altered by partially spanning logjams?
- What relative logjam spanning ratios are required to sufficiently increase flow heterogeneity?
- How can the backwater rise due to partially spanning logjams be estimated to inform both restoration efforts and flood hazard assessments?
This project is a collaboration with Prof. Heidi Nepf at the Massachusetts Institute of Technology and Dr. Elizabeth Follett at Cardiff University.
This project has gratefully received funding from the European Union’s Horizon 2020 research and innovation programme, in form of a Marie Skłodowska-Curie Individual Fellowship (MSCA-IF), under grant agreement number 887254. Furthermore, the SmartWood_3D project is supported by the Swiss Federal Office for the Environment (BAFU), which is highly appreciated.
Large wood (LW) plays an important role in fluvial systems as it moderates stream power, regulates sediment transport and provides habitat for fish and other organisms. Besides the beneficial effects of a balanced wood budget in rivers, challenges arise for abundant quantities of accessible LW. Large quantities of wood show negative effects on stream ecology, river-crossing infrastructure and flood mitigation. The sudden and disastrous occurrence of LW during floods regularly affects communities and stream systems all over the world.
Due to a lack of applicable methodologies in LW research, little is known about transport dynamics and depositional behaviour of wood in rivers to date. In order to expand the current understanding of flow-sediment-wood interaction processes, especially at higher flow rates, specific and profound research is required. Therefore, novel methods, applying innovative technologies, are developed and verified at VAW.
For the first time, nine-degree of freedom (9-DoF) smart sensors with external GPS units are implanted into prototype logs – SmartWood (Figure 1) – for exploring LW movement dynamics in the field. The sensors comprise of an accelerometer, gyroscope, and magnetometer, and allow the record of mobilisation, transport, and depositional processes at high resolution. Gained sensor data will allow for the reconstruction of 3D trajectories of LW transport as well as provide insights into LW impact forces, arising from collisions with in-stream structures and channel boundaries.
Often, transported wood pieces are entrapped by in-stream structures (e.g., bridge piers and weirs), and triggering the formation of LW accumulations. These accumulations significantly affect flow hydraulics and increases stress on the obstructed structure, yet little is known about accumulation characteristics such as accumulation volume, packing arrangement or porosity, which are of great interest for the prediction of backwater and resulting effects on channel morphology. An image-based methodology - Structure from Motion (SfM) photogrammetry – offers promising perspectives and new insights into LW accumulations and is currently applied by this SmartWood_3D project. SfM Photogrammetry generates 3D point cloud and mesh models from 2D images (Figure 2), which are of use for the computation of accumulation volume and packing arrangement. An efficient workflow-pipeline is elaborated and will allow for rapid 3D model generation and assessment of prototype LW accumulations.
SmartWood_3D aims:
- to introduce state-of-the-art technologies into LW research, in order to advance applicable methodologies and to provide an efficient workflow-pipeline,
- to quantify LW movement processes, in particular LW mobilisation, transport, and deposition,
- to determine impact forces, which arise from collisions of LW with river-crossing infrastructures and channel boundaries,
- to generate accurate 3D models of random LW jams in the field for a range of applications, such as obtaining information about accumulation volume, packing arrangement, or porosity, and
- to merge the gained results and improve the understanding of complex flow-sediment-wood interaction processes in rivers.
An improved understanding of LW movement processes is essential to better predict arising impacts on channel morphology and river-crossing infrastructure. Gained results will contribute in a more reliable risk assessment for wood prone stream systems, advanced river and forestry management strategies under consideration of LW conveyance and filtering of critical key-logs, and to realize a concept that allows wood in rivers, as a natural and environmentally important element.
Update: August 2022
Ten custom designed smart sensor units were purchased and installed into 4 m long prototype wood logs with a mean diameter of 0.33 m, so called SmartWood. During the floods in summer 2021, SmartWood experiments were carried out for the first time at field-scale, considering three Swiss rivers; the Grosse Melchaa River in the canton of Obwalden, as well as the Thur and Limmat River in the canton of Zurich (Figure 3). The sensors and SmartWood-logs met the requirements for field-scale experiments, while novel data about LW dynamics could be collected. These data are currently processed and analysed.
The gained sensor data will allow for the assessment of LW orientation and transport velocity from inertial measurements (acceleration, angular velocity, magnetic flux). In addition, they also show great potential for the reconstruction of 3D transport trajectories (from IMU and GPS data). Furthermore, SmartWood data provide new insights into LW impact forces, arising from collisions with in-stream structures and channel boundaries.
Besides focusing on LW movement dynamics in Swiss rivers, characteristics (e.g., volume, porosity, structural alignment) of prototype LW accumulations were assessed in the field (Figure 4). Intact LW accumulations all across Switzerland were mapped via close-range aerial Structure-from-Motion photogrammetry, yielding 3D replicates (digital twin models) of the examined wooden deposit and thus allowing precise volumetric measurements (2.5D and 3D volume). Afterwards, the deposits were removed from the stream system and thoroughly quantified (weighed, processing to wood chips), while wooden samples were brought to the laboratory for detailed analysis (wet density, dry density, water content). Based on the obtained data and parameter, the net wood volume (solid cubic meter, solid wood volume) is estimated. The solid wood volume represents an important measure besides the 2.5D and 3D volume (bulk volumes) for porosity estimates. Currently, an efficient workflow-pipeline is elaborated, which will enable the rapid assessment of prototype LW accumulations using digital twin models.
Data gained in the course of the SmartWood_3D project are of relevance for river managers and engineers to maintain existing as well as design new in-stream infrastructure. Furthermore, the generated data provide invaluable information about transport and depositional dynamics which are required for the calibration and verification in numerical modelling.
We appreciate the support received from power plant operators (EKZ, EWO, KHR, Axpo, Ofible), the local community Sarnen, the Building Department of the Canton of Zurich as well as the Swiss Federal Office for the Environment (BAFU) and every single person involved.
Many traditional bed stabilization measures, e.g. concrete check dams, are present in steep mountain streams in the Swiss Alps. These measures reduce bed and bank erosion, regulate sediment transport and therefore mitigate flood protection issues. However, these measures are associated with an unnatural bed morphology, high construction and maintenance costs, and they tend to have an abrupt failing mechanism in case of overload.
In consideration of the actual Water Protection Act, measures with a more natural bed morphology are required. Step-pool sequences (Fig. 1) have a high potential to replace the traditional, unnatural bed stabilization measures. Attempts of using self-stabilizing step-pool sequences for bed stabilization in steep mountain streams already exist. Coarse material is added to the streambed and step-pool sequences are formed during high flood events with recurrence interval of many decades. Consequently, a long time period is required for the formation of the step-pool sequences and the process is associated with high material losses that might decrease flood safety in low land rivers.
Artificial step-pool sequences are a potential solution to the above mentioned problems. Step-pool sequences are artificially built and thus the stabilizing effect is immediately present after the installation without an extensive material loss. Nevertheless, no design criteria for artificial step-pool sequences are available yet.
The main scope of this study is to investigate the effect of bed slope, various step-pool configurations (block size, block arrangement, counter steps, step length), channel width, sediment transport, and bank geometry on stability and failure mechanisms of step-pool sequences. The overall aim is to develop design criteria for artificial step-pool sequences that meet the following needs: Efficiency with regard to construction and maintenance costs, a more natural morphology than conventional bed stabilization measures and a more gradual failure mechanism in case of overload compared to the traditional bed stabilization measures.
The project is financed by the Federal Office for the Environment (FOEN) and supported by an expert group.
Natural river systems are shaped by dynamic processes such as water flow, sediment, and wood transport and characterized by high hydro-geomorphological variability. Riverine ecosystems are small-scale, intricate mosaics of aquatic and terrestrial habitats and support high biodiversity. However, numerous rivers in Switzerland and worldwide are heavily impacted by human interference, for example, by channelization (i.e., artificial confinement to a narrow riverbed designed for efficient water and sediment conveyance). In addition, the sediment continuum of rivers is interrupted by transversal structures such as weirs or sediment retention basins. The combined effect of sediment deficit and channelization transformed dynamic alluvial river systems with extensive floodplains into incised, straight rivers with a flat riverbed and little morphodynamic activity. The resulting uniform and static river systems cannot adequately sustain riverine flora and fauna adapted to the high spatiotemporal dynamics of natural river systems. Instead, the flow field is homogeneous, the riverbed is coarse and armored, and connectivity between aquatic and terrestrial habitats is interrupted.
Modern river restoration works towards reestablishing dynamic riverine processes such as sediment transport and the associated morphodynamic activity. Dynamic river widenings are river reaches of a few hundred meters to several kilometers in length designated to increase local river width and morphological variability (Figure 1). After removing bank protection and implementing further initiation measures, bank erosion may increase channel width and sediment input to the river, and a more near-natural morphology may reappear.
The development of dynamic river widenings may be determined by many factors, for example, floodplain erodibility, vegetation growth, or longitudinal slope. Previous studies conducted experiments with a longitudinal slope of approximately one percent to investigate the morphodynamic adjustments of reach scale river widenings and the influence of sediment supply from upstream. In this study, we continue this research and investigate how the widening process is developing for a longitudinal slope of approximately 0.3 percent. The main questions are:
- Does sediment supply influence the morphological development of one-sided dynamic river widenings with low channel slope?
- What are different effects for the morphology investigating implementations of local flow deceleration (e.g. ELJ, initial widening, …) and flow acceleration measures (e.g. flow deflector, …)?
- How does dynamic river widening effect the availability of flood and drought refugia?
These questions will be answered based on a large-scale mobile-bed laboratory experiments representing a one-sided dynamic river widening in a gravel-bed river at VAW. The laboratory results are complemented with numerical modeling using BASEMENT.
This study is part of the transdisciplinary research project Resilient Rivers: Refugia – Connectivity – Stepping stones in the research program Hydraulic structures and ecology (2022-2026) and is co-financed by the Swiss Federal Office for the Environment (FOEN).
Investigation of pier scour protection measures in a fine sandy subsoil on the River Aare
Over the last few years, a pronounced scour of approx. 10 m has developed in the Aare between two 35 m long and 6 m wide pillar foundations of the highway bridge near Wangen an der Aare. Due to the planned expansion of the lanes, two additional piers with foundations (approx. 20m long and 6.5m wide) are to be built in front of the existing piers. Due to the pronounced scour hole, it is difficult to estimate the changes caused by the construction in terms of hydraulics and the loads on the bed.
In addition, the bed of the Aare consists of an unusual two-layer structure: a thick layer of fine sand (dm ≈ 0.1 - 0.2 mm) is covered by a thin top layer of coarser bedload and protected from scour formation. However, as soon as the top layer breaks up, this can lead to accelerated erosion due to the fine sands, which in turn can endanger the pier foundations or bank protection.
The following issues are to be investigated in the 1:35 hydraulic model using a two-layer structure of the subsoil:
- Study of variants and optimization of scour protection measures
- Construction conditions
- Robustness of the best variant: testing in the EHQ and with alluvial wood
In addition to bed measurements using laser scanning and water level measurements, velocity measurements are also carried out using acoustic Doppler velocimeters (ADV).