Dr. Stuart N. Thomson
Research Scientist
Department of Geosciences
University of Arizona

Project Title

Work related to this project was the lead cover story of the September 16th, 2010 issue of Nature

Thomson, S.N., Brandon, M.T., Reiners, P.W., Tomkin, J.H., Vásquez, C. & Wilson, N.J. (2010). Glaciation as a destructive and constructive control on mountain building. Nature, 467, p. 313-317, doi: 10.1038/nature09365.

In the News:
Science Now, September 15th, 2010
MSNBC, September 15th, 2010
Wired Science, September 15th, 2010
Scientific American, September 15th, 2010
Christian Science Monitor, September 15th, 2010
University of Arizona News, September 15th, 2010
New Scientist, September 15th, 2010
ORF (Austrian Broadcasting Corporation), September 15th, 2010 (in German)
Der Spiegel, September 16th, 2010 (in German)
Arizona Daily Star, September 18th, 2010

Project Summary

The remarkable match between the glacial equilibrium line altitude (ELA) and summit elevations in many active mountain ranges of the world has led to the supposition that glaciers act as an erosional buzzsaw, whereby high glacial erosion rates act to effectively remove any topography tectonically uplifted through the ELA.

However, evidence for "the glacial buzzsaw" is based primarily on empirical or observational evidence. Only preliminary attempts have been made to theoretically simulate predictions of this process using realistic quantitative models of glacial erosion.

Understanding the importance of glacial erosion is complicated by the fact that elevated topographies are often the result of plate convergence, and so an appreciation of the feedbacks involved between orogenic wedge mechanics and surface processes is required. A conceptual model incorporating wedge dynamics predicts that the temporal and spatial patterns of uplift and erosion for both a weak and strong glacial buzzsaw in an active orogen contrast significantly with erosion patterns of an orogen that responds to the buzzsaw by passive isostasy. To test these model predictions, and hence better judge whether glacial erosion acts as a 'strong' or 'weak' buzzsaw (i.e. how efficient is glacial erosion?) we are applying combined apatite (U-Th)/He and fission-track low temperature thermochronology to evaluate changes in erosion rates and spatial erosion patterns following the onset of late Cenozoic glaciation in the Patagonian Andes.

Why the Patagonian Andes?

The Patagonian Andes are the orogen where the match between the height of the ELA and summit elevation was first recognized. They represent an exceptional natural laboratory. As well as having a suitable and well-documented tectonic and climatic history, their north-south range also provides an opportunity to study a spatially varying glacial history.

Figure 1. N-S Profile showing the excellent match between snowline and summit elevation in the Patagonian Andes (by Steve Porter, as presented in Broecker and Denton, 1990)


Empirical evidence for a glacial buzzsaw in Patagonia

The Liquiñe-Ofqui fault zone at about 44° to 45°S comprises several fault blocks that have undergone quite distinct late Cenozoic differential uplift and erosion as a result of oblique-slip offset along several major faults (Thomson, 2002) - see Figure 2 below. Despite this the topography of each of these fault blocks is remarkably similar. The hypometry of each of these blocks show that the mean and modal elevation is close to that of the long term mean elevation of the ELA between the last glacial maximim and today (Figure 3). This is consistent with similar findings in the NW Himalaya (Brozovic et al., 1997) where is was stated that "The long-term [mean late Cenozoic] ELA forms an upper envelope on the development of topography through which only a relatively small amount of [rock] material is allowed to pass".

Figure 2. (a) Fission-track data location map; (b) Estimates of total denudation and denudation rates estimated from the fission-track data; (c) SRTM DEM demonstrating uniformity of topography across all fault blocks; (d) Geological map, demonstrating the dominance of granitoid bedrock in all fault blocks, thus eliminating differential rock strength as a possible cause of differential denudation.

LOFZ Plots

Figure 3. Hypsometry for each of the four fault blocks illustrated in Figure 2. Colours refer to different fault blocks in Figure 2(b). In black is the hypsometry of all the fault blocks combined.

LOFZ Plots2

Conceptual Model

What should be the expected reponse of an orogen to the onset of glacial buzzsaw-type erosion
(i.e. the lowering of the ELA)?

We specify four end-member scenarios:
1) Is the glacial buzzsaw (or efficiency of glacial erosion) (a) strong or (b) weak / non-existent?

2) Does the orogen respond to glacial buzzsaw erosion by (a) passive isostatic rebound or (b) by active uplift driven by internal deformation of a critical orogenic wedge?

Weak Buzzsaw, Passive Response (1b + 2a) and Weak Buzzsaw, Active Response (1a + 2b)

Figure 4. If the buzzsaw effect is weak, regional scale erosion rates will not be significantly altered from their background rate by lowering the ELA.

Strong Buzzsaw, Passive Response (1a + 2b)

Figure 5. If the topography above the ELA is rapidly lowered, isostatic compensation will increase erosion rates in the central highest part of the range by ca. 5x the rate of ELA lowering.

Strong Buzzsaw, Active Response (1a + 2a)

Figure 6. If the topography above the ELA is rapidly lowered, erosion is more strongly concentrated at the crest of the range (in comparison to the passive case). To maintain a critical taper, the buzzsaw induced reduction in height of the orogen, also results in the orogen being reduced in width.

Thermochronologic Results

Figure 1. Topographic, and tectonic map of southernmost South America showing sample locations.
Shown are location of samples used to plot age-elevation relationships in Figure 2, and position of transects shown in Figures 3 and 4.

Figure 2. Age-elevation relationships from two high relief transects.
(a) Apatite fission track age-elevation relationship (AER) from 38-40°S.
(b) Apatite (U-Th)/He AER from same samples showing younger ages at lower elevations indicative of an onset of enhanced erosion at ca. 7 Ma coeval with the onset of major glaciation in southern South America.
(c) Apatite fission-track AER from 43-44°S showing a similar transition to younger ages with longer mean track length distributions (histograms shown with mean track length) at lower elevations at ca. 7 Ma. The break-in slope approximates the former closure depth of the thermochronometric system (ca. 4.1 km for AFT for erosion rates of ca. 0.5 mm/yr) immediately prior to accelerated erosion at ca. 7 Ma. This equates to a time-averaged erosion rate of ca. 0.6 mm/yr since this time. The samples close to sea-level represent an additional 1.5 km of erosion (total of 5.6km erosion) requiring a time-averaged erosion rate of 0.8mm/yr since 7 Ma.
(d) Apatite (U-Th)/He ages from the same samples show a similar AER.
All age error bars where larger than symbol are 1 sigma.

Figure 3. Four east-west transects across the Patagonian Andes at different latitudes.
(a-c) The three northernmost transects show a distinct 'u-shaped' age minima for both thermochronometers (less than ca. 5 Ma) on the windward (enhanced precipitation) side of the range (new AFT ages in blue, new AHe ages in green and previously published AFT ages in light blue).
(d) The profile furthest south shows no such relationship, despite the presence of widespread fiords and ongoing glaciation. Transects include age data, mean, maximum, and minimum elevation information from a 1° swath (a, c, d) or 1/2° swath (b) either side of each transect. Mean annual precipitation data from the UNEP/GRID database (http://www.grid.unep.ch/). The highlighted region of Miocene magmatism places an upper age limit on most of the samples from this part of the transect.

Figure 4. Latitudinal swath profiles showing apatite (U-Th)/He (a) and apatite fission track ages (b).
Also plotted are maximum and mean elevation (from a 4° E-W swath centred along the profile in Figure 1), as well as the modern and mean glacial ELA. All age error bars are 1 sigma. Note that with the exception of some ages from independently dated Pliocene (5 Ma) dacite at ca. 50°S, the minimum AHe and AFT ages both show a distinct increase south of about 45°S implying a significant decrease in total amounts and rates of post 10 Ma erosion at these latitudes despite dominantly glacial conditions here since ca. 7 Ma.

Figure 5. 3D representation of modern and LGM ELA compared to mean (10km radius) and 30 second resolution SRTM 30 plus topography. (ELA data from Aniya (1988). Arctic and Alpine Research, v. 20, p. 179-187; Aniya et al. (1997). Arctic and Alpine Research, v. 29, p. 1-12; Broecker & Denton (1989). Geochimica et Cosmochimica Acta, v. 53, p. 2465-2501; Porter (1981). Quaternary Research, v. 16, p. 263-292; Hubbard et al. (2005). Geografiska Annaler, v. 87, p. 375-391; and the NSIDC World Glacier Inventory http://nsidc.org/data/glacier_inventory/ ). This figure demonstrates the substantial portion of the southernmost Patagonian Andes elevation above the ELA, despite predominantly glacial conditions here since at least the late Miocene.

Figure 6. Cartoon of critical taper orogen (modeled on Patagonian Andes) showing how glacial protection (and hence a slow down or cessation in significant bedrock glacial erosion) has a contructive control on mountain belt development, whereby if accretion rates remain constant (as has been the case in the southernmost Patagonian Andes since at least 10 Ma), then to maintain its critical taper angle, the orogen is forced to grow in both height and width

This material is based upon work supported by the National Science Foundation under Grant No. 0447140. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

Last Modified: September 27th, 2011