Cryospheric Sciences

Debris-covered glaciers

Image of the Week – When “Ice, Ice Baby” puts rocks “Under Pressure”

Image 1: Composite image of the Aiguille Verte, the heavily-fractured headwall of the Glacier d’Argentière near Chamonix, France [Credit: D. Dennis].

Bowie and Queen said it first, and Vanilla Ice brought it back. But now, I’ve set out to quantify it: Pressure. Rocks in glacial landscapes can experience many different kinds of pressure (forces), from sources like regional tectonics or even the weight of the glacier itself. Our hypothesis is that smaller-scale pressures, caused by the formation of ice in small bedrock cracks (frost-weathering), have a large effect on the sculpting of landscapes in cold regions. This post will share how we evaluate these processes and their dependence on temperature, as well as discussing the broader effects for glacier and glacial landscape evolution.

Walking through the valley in the shadow of glaciers

Growing up just outside Glacier National Park, USA, at nearly the exact edge of the former Laurentide Ice Sheet, I became familiar with the romantic lore of how we understand glacial landscapes (Images 2, 3). Observing these glacial landscapes later throughout my formal Earth science education, I came to understand mountains as passive resistors to the relentless efficiency of glacier advance, erosion, and retreat—offering evidence of past glaciations but nonetheless devoid of agency in the rise and fall of icy stadials.

My current PhD research, however, investigates a slightly-modified premise: that glaciers and their landscapes respond in concert with climate, and that dividing the dynamics governing the ice and the rock may not be as straightforward as once thought. My work is a sub-project of the Climate Sensitivity of Glacial Landscape Dynamics (COLD) project, funded by the European Research Council (ERC) and lead by Dirk Scherler at the Deutches GeoForschungsZentrum (GFZ) in Potsdam, Germany.

Image 2: The author on holiday in Glacier National Park, Montana, circa 2001, demonstrating an early aptitude for glacial geomorphology and cosmogenic nuclide geochemistry. His affinity for popular German footwear at a young age foreshadowed his eventual move to Germany to study glaciology and geomorphology [Credit: D. Dennis].

Image 3: This image of Chamonix Valley and the M. Blanc massif conceptually outlines how average annual temperature may change with elevation in steep hillslopes. The highest peaks in the massif tower up to nearly 4000 m over Chamonix Valley, which sits at appx.1000m. This corresponds to a nearly ~20 °C difference in annual average temperature. [Image adapted from Google Earth].

Temperature as a control in glacial landscapes

Glaciers exist in locations with temperatures that are, for some portion of the year, below freezing, as this is a condition required for snow to persist through the melt season and to form ice. Temperature is therefore an important primary control on the stability of glaciers. These cold temperatures, however, impact mountain environments beyond just the formation/decline of glaciers, and several decades of recent research have shown that temperature is an important controlling factor on the type and magnitude of erosion (the act of dislodging and transporting rock) in cold landscapes.

Mountain glacier valleys are commonly characterized by steep head- and sidewalls which frame the glacier within (like in our Image of the Week). At our field sites in the French, Swiss, and Italian Alps, these rockwalls can tower up to 1500 m above the surface of the glaciers, corresponding to a temperature gradient of ~10 degrees (Image 3). Therefore, the rocks at different elevations are exposed to different temperature conditions, which could lead to differences in the rate of erosion.

Image 4: Permafrost degradation and frost-weathering in the steep hillslopes of the M. Blanc massif commonly lead to the deposition of debris on the glaciers at the base of the mountains. Shown here is Glacier d’Argentière (France) with patches of surface debris [Credit: D. Dennis].

Erosion in steep rockwall faces

Frost-weathering processes occur only at temperatures at or below zero, therefore requiring the same cold temperature conditions that form glaciers. At these temperatures, liquid water present in small cracks in the bedrock freezes. The pressure exerted on the rock by the ice as it freezes causes the rock to fracture, leading to large cracks in the bedrock (Image 5). Erosion occurs when the ice in the crack becomes large enough and its corresponding fracture wide enough that the rock can no longer remain attached and it falls from the rockwall surface.

Erosion can also occur when the ice in the crack melts and no longer “cements” the surface together. Because temperatures in glacial landscapes are commonly quite cold, much of the bedrock is considered permafrost (permanently-frozen ground), and remains frozen throughout the year. In the Alps, however, warmer temperatures over the past decades have caused the permafrost to thaw, melting the lenses of ice and causing larger and more frequent rockfalls.

Temperature conditions are therefore important for both the rate at which cracks form in rocks (and erode from the surface) in addition to permafrost stability and the size/frequency of rockfalls. As temperatures change in mountain regions due to global warming, this could lead to considerable changes in debris production.

Image 5: A cropped version of our Image of the Week, showing the base of the Aiguille Verte, headwall of Glacier d’Argentière. Large fractures in the bedrock are clearly visible. These may have grown from much smaller cracks that formed due to frost-weathering.

The hillslope/glacier surface connection

After material erodes from the surface of the headwall, it is often deposited onto the surface of the glacier (Image 3). As mentioned above, the deposition of material can occur both at a constant rate or sporadically (as in the case of permafrost-thaw rockfalls), depending on the controlling process. As such, determining the actual representative rate at which these headwalls erode is challenging.

Though this work can be complicated, we believe it to be important, as debris deposited on the surface of glaciers can insulate the ice from the effects of temperature (Image 4, Video 1). Though the global distribution of debris-covered glaciers is much smaller than debris-free glaciers, debris-covered glaciers make up a non-trivial fraction of the glaciers in populated mountain regions where they may be important fresh water sources, contribute to glacial hazards, or allow for the generation of hydropower. Understanding the supply of debris to these glaciers (via erosion), and how it may change, is therefore an important component of forecasting their evolution under warming climates.

Video 1: This drone footage from the Arolla Glacier, Switzerland, shows the steep relief which can develop as a result of differential melting. Debris thicker than 2-4 cm insulate ice, leading to topographic relief on the glacier surface as exposed ice melts and covered ice is protected. [Credit: D. Gök, GFZ]

Re-evaluating the dynamic glacial landscape

Though studies of frost-cracking and debris-covered glaciers individually are not necessarily brand new inventions, our methods for combining the two are rather novel. In doing so, we are linking the evolution of glacier with the evolution of the landscape itself, and investigating an interesting feedback loop induced by changes in climate. Should erosion rates increase with warmer temperatures, and the mountains therefore supply more debris to glacier surfaces, this could extend the “lifetime” of the glacier by insulating it; likewise, if erosion rates decrease, less debris supplied to already-covered glaciers could lead to less insulation and (comparatively) higher melt rates. This interplay demonstrates the complexity of Earth system processes, and the need to take these complexities into account when considering the effects of climate changes.

To summarize

Pressure, pushing down on rock,
Pushing laterally against rock, can cause them to fall.
Under (thick) debris, glacier melt will slow,
Despite higher temperatures,
And global warming.

Will it ever stop? I don’t know.
Turn up the temperatures, then no more (ice and) snow.
At the end of the day, frost-weathering needs ice,
When water can’t freeze, ice-cracking’s no dice.

Edited by David Docquier

Donovan Dennis is a PhD student at the Deutches GeoForschungsZentrum in Potsdam, Germany. He is interested in many aspects of glaciology and glacial geomorphology, and currently investigates the geomorphic feedbacks on glacial landscape erosion. He previously worked on post-deposition alteration of stable water isotope signals in snow and ice. He tweets as @donovan__dennis.

Contact Email:


Image of the Week – We walked the Talk to Everest

Fig. 1: Group photo with Mount Everest backdrop following presentations at the Sagarmatha National Park office in Namche Bazar (3,500 m a.s.l) with 60 participants (wrapped up against the cold temperatures). [Credit: Dhananjay Regmi].

The 12 day “Walk the Talk” Field Conference and Community Consultation through Sagarmatha National Park, Nepal, discussed a wide range of research outputs with local communities, tourists, and officials. Topics covered glaciers, mountains, environmental and landscape change, Sherpa livelihoods, tourism, and natural hazards. The conference, organised by Himalayan Research Expeditions, was the first of its kind, designed to receive community input into research topics and pursue applied benefits. Scott and Katie were two of the participants, presenting work from their PhDs in the Everest region and the NERC-funded EverDrill project.

Presentations and discussions

The team of international and Nepali scientists gave presentations every evening, trekking each day between six different villages along the Everest Base Camp trail. We were also joined by officials from the Nepal Department of Tourism and the Mountain Institute. The highest destination for the conference was Imja Glacial Lake, at over 5,000 m elevation, where we viewed first-hand the results of a recent $7 million project to lower the lake water level, aiming to reduce the risk of an outburst flood.

The Sagarmatha National Park has been a focus for scientists of many disciplines for decades. As well as thousands of tourists trekking to Everest Base Camp each year, it is also frequented by those hoping to summit Mount Everest (Sagarmatha). The park has therefore experienced significant change over a relatively short timescale as it copes with this huge influx of people. Presentations for the “Walk the Talk” conference ranged from impacts of tourism (for example, on local people, yak breeding and waste disposal) to natural hazards such as glacial lake outburst floods and landslides.

Katie presented ongoing work from her PhD and the “EverDrill” project (Fig. 2), for which she has conducted several field seasons on Khumbu Glacier in the Sagarmatha National Park. Fieldwork has included hot-water drilling of boreholes into the glacier and installing sensors to measure ice temperature at various depths to investigate the glacier’s thermal regime. She discussed how these measurements showed that Khumbu’s ice is warmer than expected, potentially putting the glacier at risk of more rapid melting as air temperatures rise. The warmer ice towards the terminus also allows subsurface meltwater drainage, about which very little is known. Katie has also carried out fluorescent dye tracing experiments to work out how meltwater travels through Khumbu Glacier, including storage within (englacial) and on the surface (supraglacial). As Khumbu and similar glaciers retreat in the future, meltwater storage and runoff will have implications for the downstream communities who depend on such water sources.

Fig. 2: Katie presenting measurements of Khumbu Glacier’s thermal regime and hydrology at the Sagarmatha National Park headquarters in Namche Bazar (3,500 m a.s.l.). [Credit: Dhananjay Regmi].

Scott presented results from his PhD investigating melt processes and water storage on Khumbu Glacier (Fig. 3). Areas of Khumbu Glacier have thinned by up to 80 m over the last three decades and glacier flow is slowing down, which allows meltwater to pond on the glacier surface. The rugged glacier surface is pitted with ice cliffs and ponds, which act as hot-spots of melt in areas of the glacier otherwise insulated by a thick layer of rocks and sediment (debris-cover). The rapid formation, persistence, and drainage of meltwater stored on glaciers across the Himalaya is a growing concern due to the potential for outburst floods and increased rates of glacier melt. An outburst flood event that occurred in the Everest region in 2017 destroyed trekking trails and a bridge.

Fig. 3: Scott presenting a study of glacier thinning at the Sagarmatha National Park office in Namche Bazar (3,500 m a.s.l). [Credit: Dhananjay Regmi].

After the final day of trekking, an extra night was spent in the village of Lukla, before flying back to Kathmandu. Each presentation was summarised in a few slides, and collated into a full talk that was given in Nepali by Dr. Dhananjay Regmi, organiser of the conference and head of Himalayan Research Expeditions. By presenting all our research in Nepali, more local people attended and were able to hear about and suggest new directions for research in the valley. This presentation was given again two days later, also in Nepali, at the Department of Tourism in Kathmandu, for locals who had already travelled back to the city to avoid the high-elevation winter chill.

Outreach activities

Fig. 4: The projection augmented relief model shown after presentations in the village of Phortse. The inset shows glacier velocity data projected onto the glaciers in the Everest region. [Credit: Gu Changjun and Scott Watson].

We designed outreach activities and leaflets to enhance the PowerPoint presentations given at each village by providing interactive demonstrations of key research concepts and results. Scott used an AGU Celebrate 100 grant to design a projection augmented relief model (PARM) of the Everest region (Fig. 4). The PARM system projected research results including glacier velocity, mass loss, ice thickness, temperature, and animations of glacier flow, onto a 3D model, which stimulated discussion of the research. The 3D model allowed the local communities to easily visualise the data in the context of well-known mountain peaks and glaciers, and to observe the changing environment (such as the expansion of Imja Lake) from a projected time-lapse animation.

Fig. 5: Katie demonstrating glacier thermal regime and hydrology using a 3D model to conduct example dye tracing experiments. The lower panel is a GIF showing the dye tracing. [Credit: Scott Watson and Katie Miles].

Katie’s interactive outreach was to demonstrate dye tracing experiments on a 3D model of Khumbu Glacier (Fig. 5). Food colouring was used to “dye” the water, which was “injected” into a supraglacial stream, then “disappeared” into the glacier. The side view into the glacier showed this water flowing through and beneath the ice, before emerging back at the surface, flowing through surface ponds and exiting the glacier at its terminus. The side view also showed the approximate ice temperatures measured by the EverDrill project, which actively showed where (and why) the glacier is experiencing more melt.

The model was very well received by scientists and locals – while the water was being injected, we would explain what was happening in both English and Nepali, and there were always plenty of questions. While the dye tracing experiments didn’t work perfectly every time, surface floods offered an opportunity to talk about other hazards that have been recently observed on Khumbu Glacier.


The “Walk the Talk” Field Conference and Community Consultation was a new style of conference, aiming to communicate a wide range of research topics in the Everest region of Nepal and the Sagarmatha National Park. The combination of high-elevation trekking and presentations was sometimes tiring, but the trek facilitated discussions about the landscape we were immersed in and was a fantastic learning experience. It is hoped that the conference will travel to different locations in the future to share research and understand the priorities of other communities in Nepal.

Further reading

Edited by Violaine Coulon

Scott Watson is a Postdoc at the University of Arizona, USA, studying glaciers in the Everest region and the surface interactions of supraglacial ponds and ice cliffs. He also investigates natural hazards and glacial lake outburst floods. Tweets @CScottWatson. Website:




Katie Miles is a PhD student at the Centre for Glaciology, Aberystwyth University, UK, studying the internal structure and subsurface hydrology of high-elevation debris-covered glaciers in the Himalaya through borehole-based investigations and dye tracing experiments. Tweets @Katie_Miles_851. EverDrill website: