Roches moutonnées

Roches moutonnées are asymmetric bedrock bumps or hills with a gently sloping and abraded upglacier (stoss) face and a quarried (or plucked) downglacier (lee) face that is typically blunter1,2. A good example of a roche moutonnée is shown in the image below.

Roche moutonnée from near Castle Loch, southwest Scotland, with a gently sloping (abraded) stoss face and a blunt (quarried) lee face. Ice flow was from left to right. Photo: David Baird

Roches moutonnées range in size from several metres to several hundreds of metres across, and often occur in clusters1 (see image below). They may be found emerging from beneath actively deglaciating ice masses (see image below), or on the sides and bottom of deglaciated valleys where they were once overridden by glacial ice3,4. Their distinctive form, which is partly linked with the orientation of glacier flow, make roches moutonnées useful to glaciologists aiming to reconstruct the flow direction of former glaciers.

Cluster of roches moutonnées (white arrows) in Porsangerfjorden, northern Norway. Ice flow was from right to left. Photo: Arnstein Rønning
Roches moutonnées emerging from beneath Goldbergkees glacier (Austria) as the ice thins and retreats. Photo: Ewald Gabardi

How do roches mountonnées form?

Roches mountonnées develop their distinctive morphology due to the pattern of stress on a bedrock surface beneath a sliding glacier, as shown in the diagram below. On the stoss side of bedrock bumps, normal stresses are relatively high and particles embedded in the ice are moved across the underlying surface where they carry out abrasion5,6. The evidence of such abrasion is the common occurrence of striations (i.e. scores and scratches on bedrock) on the sloping upper surface and flanks of roches moutonnées (see image below).

Formation of a roche moutonnée as a result of stress differences over the bedrock surface. High normal stress (pressure) on the stoss face results in bedrock abrasion, whereas lower normal stresses (pressure) on the lee face often allow a cavity to form, which promotes quarrying of bedrock along lines of existing weakness (e.g. bedrock joints). Diagram: Jacob M. Bendle
Striations on the flank of a roche moutonnée in Mount Rainier National Park, USA, giving evidence of glacial abrasion. Photo: Walter Siegmund

On the lee side of bedrock bumps, normal stresses are lower, which allows a cavity to form between the ice and bed (see diagram above) and prevents abrasion. In its place, bed cavities increase stress build up in the bedrock immediately upstream of the cavity, causing rock fracture and erosion by quarrying (or plucking). This process is particularly efficient where water pressure at the bed regularly changes3,7,8,9 (see diagram below).

The importance of bed cavities in roche moutonnée formation. In T1, the water pressure (pw) present in the bed cavity in the lee of a bedrock bump offsets the downward directed ice overburden pressure (pi), preventing bedrock fracture. However, in T2, the water has drained, and a high stress zone (red) develops in the bedrock around the cavity edge, which causes rock fracture and quarrying (plucking) to occur. Diagram: Jacob M. Bendle

The quarrying of rock at the lee end of roches mountonnées is also strongly influenced by the joint distribution in the parent rock3, and determines the size and shape of quarried rock fragments (see diagram below).

The importance of bedrock joint structure in the evolution of quarrying of a roche moutonnée lee face. The orange lines depict the progressive upglacier migration of the lee face as bedrock fragments are progressively plucked along lines of weakness (joints). Diagram: Jacob M. Bendle

What do roches mountonnées tell us about former glaciers?

Through an understanding of how roches mountonnées are formed, glaciologists are able to make inferences about the nature of past glacier systems where such landforms are found.

As roches mountonnées are most likely to form where cavities exist at the glacier bed, it is common for them to develop where the ice overburden pressure is low (i.e. where ice is relatively thin). Such conditions occur beneath thin cirque or valley glaciers, or near the margins of ice sheets3,4,10. This also means that roches moutonnées may be more likely to develop during deglaciation, when a glacier or ice sheet thins, ice overburden pressure decreases, and gaps between the ice and bed open up11 (see diagram below).

During full glacial conditions, when ice is at its thickest, ice overburden pressure (pi) is high and the glacier presses down into bumps in the bed. As the ice thins during deglaciation, the ice overburden pressure (pi) decreases and cavities open up at the bed, promoting favourable conditions for roche moutonnée formation. Diagram: Jacob M. Bendle (based on Roberts and Long, 2005)

Because roche mountonnée formation is also aided by fluctuations in basal water pressure, they are most likely to occur beneath warm-based (temperate) glaciers with hydrological systems that direct meltwater the bed10. The fact that they contain abraded (i.e. polished and striated) surfaces (see image above) also informs glaciologists that the ice responsible for their formation was (at least at times) warm based and moving by basal sliding, as well as carrying a basal debris load.


[1] Bennett, M.R., and Glasser, N.F. (2009) Glacial Geology: Ice Sheets and Landforms. Wiley-Blackwell.

[2] Benn, D.I., and Evans, D.J.A. (2010) Glaciers and Glaciation. Routledge.

[3] Sugden, D.E., Glasser, N.F., and Clapperton, C.M. (1992) Evolution of large roches moutonnées. Geografiska Annaler, 74A, 253-264.

[4] Glasser, N.F. (2002) Scottish Landform Example 28: The large roches moutonnées of upper Deeside. Scottish Geographical Journal, 118, 129-38.

[5] Boulton, G.S. (1974) Processes and patterns of glacial erosion. In Glacial Geomorphology (ed. D.R. Coates) Springer, Dordrecht, pp. 41-87.

[6] Hallet, B. (1979) A theoretical model of glacial abrasion. Journal of Glaciology, 23, 39-50.

[7] Iverson, N.R. (1991) Potential effects of subglacial water pressure fluctuations on quarrying. Journal of Glaciology, 37, 27-36.

[8] Hallet, B. (1996) Glacial quarrying: a simple theoretical model. Annals of Glaciology, 22, 1-8.

[9] Cohen, D., Hooyer, T.S., Iverson, N.R., Thomason, J.F. and Jackson, M. (2006). Role of transient water pressure in quarrying: A subglacial experiment using acoustic emissions. Journal of Geophysical Research: Earth Surface, 111(F3).

[10] Hall, A.M., and Glasser, N.F. (2003) Reconstructing the basal thermal regime of an ice stream in a landscape of selective linear erosion: Glen Avon, Cairngorm Mountains, Scotland. Boreas, 32, 191-208.

[11] Roberts, D.H., and Long, A.J. (2005) Streamlined bedrock terrain and fast flow, Jakobshavns Isbrae, West Greenland: implications for ice stream and ice sheet dynamics. Boreas, 34, 25-42.


I am a Quaternary geologist with a focus on palaeo-ice sheet dynamics and palaeoclimate change during the last 20,000 years. I study glacial landforms to reconstruct glacier (and glacial lake) extents, dimensions and depositional processes. However, my main focus lies with the sedimentological analysis of annually-layered glacial lake sediments (known as varves) to develop continuous, high-resolution records of past ice sheet response to sub-centennial (rapid) climate shifts. Read more about me at

This site uses cookies. Find out more about this site’s cookies.