Are Miette Group Gritstones Evidence of Snowball Earth?

Did "Snowball Earth" exist?  Do Miette Group gritstones in Canada support the Snowball Earth hypothesis?

by Scott N. Stone

Introduction-- The Miette Group and Snowball Earth

The Miette Group comprises middle Neoproterozoic formations of slates, "gritstone" conglomerates, schists and dolostones that outcrop in the Canadian Rockies.  Other than the deep crystalline basement and the sedimentary formations of the Belt-Purcell Supergroup (Mesoproterozoic, 1470-1400 Ma), the Miette Group strata contain the oldest rocks in the Canadian Rockies, having formed beginning about 750-800 Ma as marine sediments falling into a newly opening ocean as the supercontinent of Rodinia rifted apart.  See Ben Gadd, Canadian Rockies Geology Road Tours (Corax Press 2008), pp. 64-67.

The Miette gritstones, which include graded turbidites with probable dropstones, are taken to be of glacial origin and Cryogenian in age.  Gadd (2008) pp. 64-67 and photo of gritstone at p. 19.  The name "Cryogenian Period" ("cryo" - cold, "gen" - producing) has been applied to the middle Neoproterozoic -- 850-630 Ma -- signifying a growing scientific consensus that this period included intervals of cold temperatures and heavy glaciations, including the Sturtian and Marinoan glaciations.  Strata in Northwest Canada deposited on the North American craton (Laurentia), including the Canadian Rockies and the Mackenzie Mountains just north of the Rockies on the border between Yukon and NW Territories, have been identified as glacial in origin despite the fact that their paleogeomagnetic signatures indicate deposition when Laurentia was located in the tropics as Rodinia first assembled then broke up, as shown in Fig.1.  See, e.g., Edith.S Day, Noel P James, Guy M Narbonne, and R.W Dalrymple, 2004, A sedimentary prelude to Marinoan glaciation, Cryogenian (Middle Neoproterozoic) Keele Formation, Mackenzie Mountains, northwestern Canada, Precambrian Research, 133 (3-4): 223-247, http://www.sciencedirect.com/science/article/pii/S0301926804001202; Francis A. Macdonald, Mark D. Schmitz, James L. Crowley, Charles F. Roots, David S. Jones,  Adam C. Maloof, Justin V. Strauss, Phoebe A. Cohen, David T. Johnston and Daniel P. Schrag, 2010, Calibrating the Cryogenian, Science, 327 (5970): 1241-1243, DOI: 10.1126/science.1183325, http://www.sciencemag.org/content/327/5970/1241.abstract; Gadd (2008) p. 65.

Fig. 1  Position of Laurentia during the Cryogenian Period.  Source: © SnowballEarth.org.

The presence of glacially-derived deposits at equatorial latitudes may support the "Snowball Earth" hypothesis, which posits that the Cryogenian included episodes in which the land and ocean surfaces of Earth were completely covered in ice, except for volcanic vents from which carbon dioxide, methane and other greenhouse gases were released.  Proponents of the Snowball Earth hypothesis, including Dr. Paul F. Hoffman, formerly of Harvard and now of the University of Victoria (British Columbia), suggest that after some millions of years of total ice cover, the buildup of volcanic greenhouse gases, which could not be absorbed by their usual sinks in the crust or oceans because of the ice barrier, would trigger warming, rapid melting of the glaciers and sea ice, and initiation of a warm interval.  Snowball Earth episodes may therefore have recurred at least twice, punctuated by rapid melting and gradual refreezing.   See http://www.snowballearth.org.

Fig. 2  Assumed mechanism by which Snowball Earth episode eventually ended after high levels of atmospheric carbon dioxide were reached..  Source: © SnowballEarth.org.

A major factor that may have created the cooling trend leading to Snowball Earth was a buildup of atmospheric oxygen beginning about 850 Ma.  Cyanobacteria in stromatolite colonies had been producing oxygen from about 2.7 Ga (see  Lepot, Kevin; Karim Benzerara, Gordon E. Brown, Pascal Philippot, 2008, Microbially influenced formation of 2.7 billion-year-old stromatolites, Nature Geoscience 1 (2): 118–21  DOI:10.1038/ngeo107), but until about 850 Ma the oxygen was depleted by reaction with dissolved ferrous (reduced) iron in seawater to created insoluble ferric (oxidized) iron that then precipitated into banded iron deposits.  Oxygen hence was not in sufficient supply to become an important constituent of the atmosphere until roughly 900-850 Ma.  See Fig. 3, taken from Wikipedia, Great Oxygenation Event.

Fig. 3.  Concentration of oxygen in the atmosphere.  Note that the current percentage of oxygen in the atmosphere is 21% and hence the current pressure of oxygen at sea level  ("PO2") is 21% times the total pressure of 1.0 atm., or 0.21 atm.  Source: Wikipedia, Great Oxygenation Event

Once excess oxygen accumulated in the atmosphere (see Fig. 3), it must have oxidized methane (a strong greenhouse gas) to carbon dioxide (a weaker greenhouse gas), leading to trapping of less heat by the atmosphere and a cooling of Earth.  Another factor causing the Earth to cool may have been the position of the supercontinent Rodinia near the equator.  See Fig. 1.  Because land surfaces reflect more solar energy than the ocean, the clustering of continents at the equator, where solar radiation is most intense, would have reduced the amount of solar energy absorbed.  In addition, tropical rainfall falling on the equatorial continents would have rapidly weathered silicate rocks into carbonates, removing atmospheric carbon dioxide in the process.  See http://www.snowballearth.org.  For a list of publication about Snowball Earth, see http://www.snowballearth.org/people.html#hoffman.

An alternative to the Snowball Earth hypothesis is the "Slushball Earth" view, in which it is accepted that much of Earth, including the tropics, was extensively glaciated, but that open areas of ocean and perhaps some unglaciated land remained.  Some computer models have shown that the positive feedback of more ice, greater albedo (reflectance), and colder temperatures may have stopped short of producing complete ice cover.  See Wikipedia, Snowball Earth.

Miette Gritstone Outcrop on Icefields Parkway

Shown at increasing levels of magnification below (Figs. 4-9) is an outcrop of Miette Group gritstone (of the Corral Creek Formation) on the west side of the Icefields Parkway (Hwy 93) 3 km due north of Lake Louise, Alberta, Canada and 1.2 road km north of the beginning of the ramp to Hwy 93 from Hwy 1, the Trans-Canada Highway.  The first three photos shown are from Google Maps, and can be viewed on Google Maps using Street View.  (The resolution of images on Google Earth for this area is not as good as on Google Maps.)  The fourth and fifth pictures (Figs. 8 & 9) were taken on July 16, 2012 during a field geology class on the Canadian Rockes led by Professors Callan Bentley of Northern Virginia Community College and Peter Berquist of Thomas Nelson Community College. An additional picture of this outcrop may be found at Gadd (2008), p. 291.

The outcrop is light grey conglomerate (called "gritstone" by British and Canadian geologists), consisting of graded and massive beds of angular, subangular and subrounded gravel (1-2 cm) of predominately quartz and to a lesser extent feldspar in a matrix of sand and finer clasts.  At least two large clasts (approx. 20-25 cm) are embedded in the gravel. The sediments above and below the large clasts are gravel of uniform size and do not display truncation of the layers below nor draping of the layers above. 

These rocks appear most consistent with deposition of glacial-derived debris as a turbidite.  That is, one can suppose that glaciers flowed into the sea and that melting of the bottom of the glaciers initially deposited a delta of glacial till consisting of poorly sorted clasts of all sizes, including abundant rock flour.  (Some sorting of the debris may have occurred if the distance between the bottom of the glacier and the sea floor was significant, since larger clasts would tend to drop most quickly through the water, with rock flour remaining suspended much longer.)  Before too long, glaciers flowing beyond the initial shoreline would remain grounded on the debris previously deposited.  The beds of debris would, depending on local topography, continue to thicken and broaden, and steeply sloping foreset beds would be pushed farther and farther into deeper water.  Underwater landslides (turbidity flows) would be triggered either by storms dislodging sediments from the steep foreset beds, or by continued glacial flows bulldozing previously deposited debris down the foreset beds.  During the turbidity flows, sorting by sediment size would occur, with coarser sediments tending to settle first (lowest) and finer sediments last (at the top of the graded bed). 

The fact that some of the rock in this outcrop appears "massive" (i.e., contains gravel of consistent size over thicknesses measured in meters) appears consistent with the observation of Ben Gadd, one of the foremost authorities on the Canadian Rockies, that Miette Gritstone turbidite beds can be tens of meters thick.  Hence it would not be surprising not to find much difference in clast size over a scale of only a few meters.  See Gadd (2008) p. 65.

It should be noted that it is not inconsistent with the Snowball Earth hypothesis to assume the existence of glaciers pushing out into the ice-covered ocean.  In both Antarctica and Greenland today, ice-stream glaciers move through both ice sheets and sea ice.  See Hoffman, P.F., 2005. On Cryogenian (Neoproterozoic) ice-sheet dynamics and the limitations of the glacial sedimentary record (28th Alex. Du Toit Memorial Lecture). Journal of South African Geology 108, 557-586.  Available online at http://www.snowballearth.org/pdf/Hoffman%28DuToit%29.pdf.

A remaining question is how can we account for the presence of the much larger clasts?  Are they dropstones, i.e., clasts embedded in glaciers that melt out and drop from the glaciers directly into the sediments we are seeing preserved in this rock?  Classic dropstones are confirmed by truncation (downward plastic deformation of layers below the dropstones) and draping (folding of sediments above the large clast to match the shape of the clast).  The large clasts in this outcrop do not fit this description.  This is not surprising.  Based on common sense, we would not expect a stone of the size in these sediments (20-25 cm long) dropping slowly through water onto gravel of the size in this outcrop (1-2 cm) to crush the gravel.  Rather, we would expect the large clast to remain sitting on top of the gravel.  But this large clast was not sitting on top of the gravel -- how would we explain why it is surrounded on all sides by undeformed gravel of uniform size?  One explanation is that this is a dropstone that dropped just as the turbidity current was coming to a stop below it.  Another explanation is that during the turbidity flow, a few large clasts from the glacial deltaic deposits were swept along with the massive amounts of gravel, rolling atop the moderately rounded gravel.  Having large rocks sliding or rolling on top of smaller rocks below is sometimes observed in landslides or, e.g., when banks of glacial till erode on roadsides.

Ultimately, because the type of glacially derived turbidite deposits seen here could occur regardless of whether the sea surface were frozen over, they do not necessarily require a Snowball Earth and thus do not necessarily support the Snowball Earth hypothesis.  The presence of "cap carbonates" signalling the melting of land and sea ice after long accumulation of carbon dioxide above ice-sealed land and oceans would be more definitive evidence of snowball Earth, but it is not seen in the outcrop discussed here.

Comments and further posts are welcomed, especially those offering further thoughts about how this outcrop may or may not tend to confirm the Snowball Earth hypothesis.  Additional field observations by those with access to this outcrop would be especially welcome.

Fig. 4. Miette Gritstone on west side of Icefields Parkway, Hwy 93 north of Lake Louise, Alberta, Canada (1.5 km north of intersection of Hwy 93 with Hwy 1, Trans-Canada Highway.  Lat. 51 deg 26' 56.24" N, Long. 116 deg 12' 54.11" W)

Fig. 5.  Miette Gritstone on west side of Icefields Parkway, Hwy 93 north of Lake Louise, Alberta, Canada (1.5 km north of intersection of Hwy 93 with Hwy 1, Trans-Canada Highway.  Lat. 51 deg 26' 56.24" N, Long. 116 deg 12' 54.11" W)

Fig. 6.  Miette Gritstone on west side of Icefields Parkway, Hwy 93 north of Lake Louise, Alberta, Canada (1.5 km north of intersection of Hwy 93 with Hwy 1, Trans-Canada Highway.  Lat. 51 deg 26' 56.24" N, Long. 116 deg 12' 54.11" W)

Fig. 7.  Miette Gritstone on west side of Icefields Parkway, Hwy 93 north of Lake Louise, Alberta, Canada (1.5 km north of intersection of Hwy 93 with Hwy 1, Trans-Canada Highway.  Lat. 51 deg 26' 56.24" N, Long. 116 deg 12' 54.11" W)

Fig. 8.  The same outcrop being examined 7/16/2012 during NVCC field geology course.  Photo credit: Cameron Perrier.

Fig. 9.  Same outcrop examined by author 7/16/2012.   Photo credit:  Patrick O'Connor.

Fig. 10  Diamictite of Pocatello Formation, Idaho, USA.  Note truncation of sediments under several large clasts, and draping over the clasts, indicating the clasts are dropstones.   Credit:  Wikipedia, Diamictite..