Rapid assessment of sediment dynamics on an active debris-flow fan

Lead Research Organisation: Durham University
Department Name: Geography


Debris flows are mixtures of sediment and water that form a primary geologic hazard in mountainous areas worldwide. Accurate assessment of debris-flow hazard depends very much on our knowledge of how debris flows transport and deposit sediment, and how they interact with, and change, their channels. Debris flows cause complex three-dimensional patterns of erosion and deposition as they move through a channel network, with important consequences for present and future flow hazards. Erosion of banks and beds can endanger infrastructure, and can significantly increase the flow volume (and thus the hazard) during an event. Conversely, widespread deposition can lead to a decrease in the capacity of the channel to transport sediment, making it more likely that later flows will abandon the channel and threaten nearby areas. For these reasons, it is important to understand both the patterns of erosion and deposition in debris-flow channels, and how they relate to the characteristics of the flows (e.g., volume, composition, velocity, and flow depth). Which types of flows produce the most bank erosion? Which flows lead to widespread deposition on the bed, thus increasing the likelihood of channel abandonment? Are there simple relationships between flow volume and the amount of channel change? We propose to address these questions by measuring very high-resolution surface topography before and after individual debris-flow events on the Illgraben fan, southwestern Switzerland. The Illgraben fan is well-suited to this because it has had 4-6 debris flows per summer since 2000; as a consequence, it is closely monitored by project partners at the Swiss Federal Institute for Forest, Snow and Landscape Research (WSL). All debris flows that cross the fan are recorded and their volume, velocity, composition, depth, and density are independently measured, making this a unique locality worldwide for studying debris flows. We will make repeated surveys of the topography using high-resolution, ground-based laser scanners, which produce three-dimensional representations of the surface with point spacings of 0.1 m along a 300 m study reach near the head of the Illgraben fan. A set of scans in spring, before the occurrence of any debris flows, will be used to define the 'starting' topography of the channel bed and banks. After passage of a major debris flow during the summer, we will re-scan the topography to create a difference map showing areas of erosion and deposition. After several different flow events, these maps will allow us to relate the type and size of flow to its effect on the channel, and will also enable us to understand how much sediment the flows lose or gain as they move across the fan. Our results will complement the monitoring efforts of the WSL on the Illgraben fan, and will be freely shared with local authorities involved in hazard assessment and mitigation.
Description The aim of this research was to understand the controls on erosion and deposition in a series of debris flows on the Illgraben fan in southwestern Switzerland. This was achieved by repeatedly scanning the Illgraben channel using a terrestrial laser scanner (TLS) before and after debris flows (of which 4-6 occur per year), producing high-resolution (20 cm) digital topographic models. From these, the spatial pattern of erosion and deposition in each flow could be determined, and we could assess the dynamic interactions between flow and channel, including the rate of downstream volume change. We had originally hypothesised that flow composition - i.e., the proportions of sediment and water, and the grain-size mixture of the sediment - should be the dominant control on the patterns of erosion in observed debris flows. To test this, we scanned a 300 m reach of the channel after 14 debris flows between 2008 and 2010, and compared the detailed patterns of erosion and deposition to the physical characteristics of the flows, including velocity, flow depth, volume, and bulk density, determined by the WSL. Calculation of surface change required development of new algorithms for treating uncertainty, which were an unexpected additional development of the project.

Our original hypothesis regarding the importance of flow composition was partly supported by the results, which showed a different downstream pattern of volume gain or loss in more coarse-grained flows as compared to muddy flows. The dominant control on erosion or deposition, however, and thus on flow volume change, proved to be flow depth, irrespective of flow composition or other factors. By stacking the results of detailed surface change estimates in multiple flows, and comparing them with measured flow depths at every point in the study reach, we were able to construct a probabilistic estimate of expected volume change for any given flow depth.

Unexpectedly, we were able to verify and extend our estimates of the rate of downstream volume change to the scale of the entire fan. This was possible because the WSL were able to provide data on flow depth and onset time from observation stations at the upstream and downstream ends of the Illgraben fan. These data allowed calculation of flow hydrographs, which could be integrated to yield estimates of debris-flow volume for flows as they entered the fan and as they exited the downstream end into the Rhone River. The results illustrated two key points: (1) the rates of volume change estimated over our 300 m study reach were congruent with rates of volume change over the entire 2 km channel length on the Illgraben fan, validating our results and allowing our findings to be scaled up; and (2) flow volumes entering the fan were a poor predictor of flow volumes exiting the fan. This latter result is critical for understanding debris-flow hazard, which is directly dependent on flow volume, and indicates a previously-unknown dynamic role for fans - they are not simply static repositories of debris-flow material, but can shape flows in unexpected ways.

The results are significant because they represent the first systematic attempt to quantify debris-flow volume change in multiple sequential events, and to relate that change to flow and channel properties. The results lend strong observational support to emerging views of entrainment by debris flows that focus on the primacy of conditions at the flow front (e.g., work by Richard Iverson, USGS, and Kimberly Hill, Univ. of Minnesota), and our data provide powerful constraints that must be obeyed by evolving numerical debris-flow models. Finally, the research represents a pioneering effort to collect near-real time measurements of surface change with TLS, and a novel application of the technique in a challenging setting (low-relief river channels). Our workflow and methodology for estimating uncertainty in surface change estimates have broad potential application for TLS in a variety of similar geomorphic settings. In sum, the objectives outlined in the original proposal were fulfilled, and the additional flow data provided by WSL allowed additional applications of the data to be explored.

A second unexpected outcome was that we were able to relate the short-term erosion and deposition observed via TLS with longer-term patterns of deposition on the fan surface observed via a high-resolution DEM. Cosmogenic radionuclide analysis of debris-flow deposits on the fan surface, funded separately, shows that flow activity on the modern fan surface extends back to about 1600 yr ago, and postdates emplacement of a large rock avalanche deposit that shifted the fan apex and the pattern of deposition. These results show that the fan is rapidly resurfaced by channel migration and avulsion over thousand-year time scales. They also show that avulsion patterns observed in analogue models of debris-flow fans are similar to those observed on field-scale fans, helping us to understand the avulsion process.
Exploitation Route The research results are highly relevant for understanding debris-flow hazard, particularly as it evolves with time or during an individual flow. The results are being used by Dr McArdell and his group to test and extend the next version of the WSL debris-flow modelling tool, RAMMS. The research has two potential routes to exploitation - through peer-reviewed outputs (see Publications), and through the collaborative involvement of the Swiss Federal Institute for Forest, Snow and Landscape Research (WSL). Collaboration with the WSL has been via Dr Brian McArdell of the Mountain Hydrology and Mass Movements Group, who was involved in all aspects of the data collection and analysis.
Sectors Environment

Description This project involved innovative measurement of erosion and deposition in debris flows, using terrestrial laser scanning. We demonstrated, for the first time, how the passage of a debris flow down a natural channel causes erosion and deposition, and how the pattern of erosion and deposition can be related to the flow depth. These results complement other studies using numerical models and controlled experiments, showing that debris flows can erode substantial amounts of sediment from the channel bed. Understanding this erosion and entrainment is important, because it increases the volume of the debris flow and thus increases the potential hazard posed by the flow as well.
First Year Of Impact 2010
Sector Environment
Description Durham University Doctoral Studentship
Amount £100,000 (GBP)
Organisation Durham University 
Sector Academic/University
Country United Kingdom
Start 03/2007 
End 11/2011
Description Institude for International Education Global Innovation Initiative
Amount $177,045 (USD)
Funding ID S-ECAGD-13-CA-149 
Organisation United States Department of State 
Sector Public
Country United States
Start 04/2015 
End 02/2018
Description NWO Rubicon Fellowship
Amount € 158,100 (EUR)
Funding ID 019.153LW.002 
Organisation NWO Rubicon Fellowship 
Sector Private
Country Netherlands
Start 09/2016 
End 08/2022
Description Swiss Federal Research Institute WSL 
Organisation Swiss Federal Institute for Forest, Snow and Landscape Research
Department WSL Institute for Snow and Avalanche Research SLF
Country Switzerland 
Sector Academic/University 
PI Contribution We provided a PhD student who undertook basic research on debris-flow processes at the Illgraben fan, Switzerland. Our team provided funding for the research (via the NERC award), equipment, and analysis.
Collaborator Contribution The partner provided access to WSL data on past debris flows, contacts with local government and civil-protection officials, access to existing instrumentation on the fan, and support in analysis and presentation of the results. The partner is also incorporating aspects of the results into other WSL outputs, notably models of debris-flow hazard.
Impact Publications
Start Year 2009