San José State University
Thayer Watkins
Silicon Valley
& Tornado Alley

Snowball Earth:
The Discovery of Evidence
That the Earth Was Once
Frozen Over from Pole to Pole


Snowball Earth refers to the contention that in the distant past the Earth froze over from pole to pole. There is considerable evidence to support this contention and the proponents now believe this freezing occurred not just once but a number of times in perhaps more than one episode of freezing and thawing. The explanation that has been developed to explain these cycles of freezing and thawing is that the weathering of silicate rocks depletes the carbon dioxide in the atmosphere resulting in a cooling of the Earth. In times when the continents are in high latitudes, as is the case now, any cooling results in the rocks being covered with snow and ice which curbs the weathering process and slows the depletion of CO2. But when the continents are massed in the low latitudes the accumulation of ice and snow in the high latitudes does not diminish the depletion of CO2 and the cooling continues until it reaches the equator. The freezing process is aided by the high degree of reflection of solar radiation by snow and ice. At the point that Earth is entirely frozen over the depletion stops.

Volcanoes however continued to add CO2 to the atmosphere. Without the depletion of CO2 from the weathering of rock CO2 accumulates. With the replenishing of the depleted CO2 the greenhouse effect then produces a warming which eventually melts the ice and snow covering of the Earth. The process once started accelerates as the uncovered land absorbs a higher proportion of the incident solar radiation. But the uncovered rocks also start depleting the CO2 in the atmosphere so the cycle could start all over again. The cycle is broken only when tectonic movements shift the continental land masses away from the equator.

As indicated above the Snowball Earth hypothesis is now a complex of hypotheses. The story has to start with the contention that the Earth once was entirely frozen.

The idea that the Earth was once frozen over first was proposed by Louis Agassiz, the person who first made humans aware of past ice ages. Agassiz was right in that there were far more extensive ice fields at times in the past than there are now. Agassiz was wrong when he contended that these more recent ice ages extended to the equator.

After Agassiz's concept of ice ages was accepted geologists found abundant evidence for numerous ice ages. One glaciation that was of interest occurred about 550 million years ago. This was a particularly interesting period in Earth's history because it was about that time that a vast diversification of life forms occurred. In the previous four and a half billion years of Earth's history life, when it existed at all, consisted exclusively of one-celled slimes. It is one of the great mysteries as to why this diversification and development of complex life forms occurred. There are a number of supported theories, but the important point is that when geologists find evidence of anything that occurred 500 to 600 million years ago alarm bells go off.

The story of Snowball Earth is largely the story of one unique individual, the geologist Paul Hoffman. Paul Hoffman was not the originator of the concept and did not even coin the name by which it is known. A British geologist, Brian Harland, was the first to contend that there was evidence for glaciers in tropical zones. Joseph Kirschvink of the California Institute of Technology formulated the theory in a remarkably fully fledged form and came up with the name Snowball Earth. Paul Hoffman personally heard of the concept Snowball Earth in a conversation with Joseph Kirschvink. But it was in Paul Hoffman's hands that the concept was raised from being an interesting speculation into a full-blown theory with supporting evidence. It took the special combination of brilliance, tenacity and obsession that Hoffman possesses to bring the concept of Snowball Earth to the attention of the geological establishment and force its testing. Paul Hoffman not only devoted his own considerable intellectual talents to the development of the Snowball Earth concept but he served as a field marshal for organizing the less disciplined genius of his Harvard colleague Dan Schrag. Hoffman when he had sufficient evidence toured the professional geological world giving presentation of the support for Snowball Earth. Hoffman's determination forced the consideration of what otherwise would have languished as a little-known, interesting idea. His efforts rankled the sensibilities of many who then devoted effort to proving Hoffman wrong. This of course was a very healthy development in the promotion of the theory. But to see the development of the theory it is necessary to start with the early evidence.

The First Glimmerings of Global Glaciation
from Brian Harland's Studies of the
Geology of Svalbard

Svalbard is a small archipelago in the remote arctic reaches of the North Atlantic. Not much has happened to Svalbard geologically over many millions of years and it provides a unique window into the very ancient past. Not many would find it all that interesting a site for field but Brian Harland, now of Cambridge University, went there in 1938 as a graduate student. After World War II he came back to carry out a program of research that culminated in the production of the definitive work on the geology of Svalbard. Harland recognized that Svalbard was not always the frigid nowhere that it is today. He found evidence that at one time Svalbard enjoyed a tropical climate. That could have been because Svalbard might once have been located in the present day tropics but tectonic movements relocated it to the Arctic. It could also have been from global climate changes that produced tropical conditions even in the Arctic. Or it could be due to a combination of both processes. Harland found that mixed in with the geologic evidence of tropical conditions in Svalbard was evidence of glaciation. The time period was just before that magical period of the burgeoning of complex life forms, the period called the Cambian Era. Harland wrote articles on what he called the Great Infra-Cambrian Glaciation.

The evidence, in part, that Harland found which indicated glacial conditions was dropstones in sedimentary rock strata. Dropstones are stones that are carried out into water bodies by icebergs. When the icebergs melt the stones that they carry drop into the sediment and become incorporated into the stone the sediment turns into.

But in strata closely associated with the layer containing dropstones there were layers of carbonate rock. Such carbonate rock formed from the precipitation of carbonates in water bodies can only form under warm conditions. Harland was not able to convince many in the geology profession of the existence of a worldwide glaciation in the Pre-Cambrian Era. In part this was because a study came out at the same time as his purporting to show that the supposed dropstones were not due to glaciers and icebergs but instead were due to underwater landslides that moved rocks into unexpected locations. It subsequently turned out that this phenomenon could not explain Harland's cases but the impact of Harland's work was minimal.

Brian Harland did pioneer the use of the magnetic field orientation of rocks to establish the general latitude of where the rocks were formed.

While the measurement of the magnetic orientation with respect to the horizontal strata is important information on the location of where a rock was formed it is not definitive because heating could have destroyed the original magnetic orientation and a new orientation acquired upon cooling.

Joseph Kirschvink Considers the Evidence
and Realizes that a Snowball Earth
Would Account For It All

Gabrielle Walker's Snowball Earth provides biographical information about the principal personalities involved in the Snowball Earth story as well as the technical issues. One individual, Joseph Kirschvink of Caltech, could well be considered a national treasure. He has worked on and solved major puzzles such as how migratory birds are able to navigate over the great distances involved in their migrations. This problem and most of the ones Joseph Kirschvink works on involve magnetism. Sometimes his expertise in magnetism leads him into investigations in related fields.

As a referee for a professional journal Kirschvink was asked to evaluate a study that used the magnetic orientation of rocks to identify the latitude of where the rocks were formed. The difficulty Kirschvink saw with study was that the original magnetic orientation of rocks might have been replaced at a later time due to heating or the diffusion of rock material into the sample rock. He started investigating the geologic evidence and did find ways to confirm that the present magnetic orientation is the same as the one imposed on the rock at formation. But in research that topic he became aware of the basic evidence, which was:

  • A stratum containing dropstones
  • An overlying stratum of carbonate rock
  • An overlying stratum of ironstone

The dropstone stratum was evidence of glacial conditions. The fact that this stratum appeared all over the world in land areas was evidence for extensive glaciation at least on land areas. The carbonate rock is formed only in warm seas. This would indicate a warming and melting of the glaciers. The melting of the glaciers is a problem since the ice fields vastly decrease the amount of solar energy the Earth would retain. It was widely believed that if the Earth froze over there would be no way it could ever thaw.

Kirschvink saw a way to explain the layer of ironstone which had been a puzzle in its own right. Before Earth's atmosphere contained oxygen a great amount of ionic iron was dissolved in Earth's oceans. This iron came from volcanoes under the oceans. When Earth acquired oxygen in the atmosphere some of it dissolved in the oceans and combined to form iron oxide, which was insoluble in water and precipitated out to form a layer of iron oxide called ironstone. This was billions of years ago and ever since no significant amount of ironstone was laid down, except in the period right after the dropstone layer. Kirschvink realized that such an ironstone layer would be created if the oceans were frozen over isolated the iron ions of the oceans from the oxygen in the air. The concentration of iron would build up until the oceans thawed, at which time the iron would precipitate out as iron oxide, ironstone.

Kirschvink also saw a way the Earth could unfreeze. The temperature of the Earth is strongly influenced by the amount of greenhouse gases in the atmosphere; i.e., water, carbon dioxide and methane. The level of carbon dioxide is the net result of its depletion through the weathering of rocks and its augmentation from volcanoes. A freezing over of the Earth would end the depletion and leave the buildup from volcano emissions.

Joseph Kirschvink discovered a mechanism that would account for the freezing and the thawing of the Earth. He packaged his creation by giving it a name, Snowball Earth. He however did not publicize it beyond publishing it in a very short article that was not widely read. It would probably have languished there if he had not mentioned it to Paul Hoffman in a brief conversation.

Joseph Kirschvink's brain is such a fountain of bold, new ideas that it is difficult for him to give much attention to any one of them. One of his later ideas is even more radical than Snowball Earth. He suggests that some of the geologic evidence can be accounted for by a massive reorientation of the Earth's crust. The Earth consists of an iron core that creates the magnetic field. On top of the core is a layer consisting of the mantle and the crust. Kirschvink proposes that this outer layer may have shifted ninety degrees with respect to the iron core. This would have put the polar zones at the equator and some areas that were on the equator would become polar. But that is another Kirschvink creation and the focus here is on Kirschvink's Snowball Earth and how Paul Hoffman and his associates developed evidence for it and refuted the challenges to it.

The Evidence from Namibia

Paul Hoffman came from Canada and pursued a career in geology as a result of his fondness for geologic field work in the Arctic. As a result of some undiplomatic remarks about his supervisor he found himself cutoff from field work in the Arctic. He chose Namibia as an alternative. The desert climate there results in exposed geological formations. To his surprise he found evidence of dropstones in some formations. He recognized that this evidence of glacial activity in a near-tropical region was significant. Plate tectonics could have accounted for the evidence. Southern Africa could have been in a polar location at the time of the rock formation. But he became aware of such evidence of glacial activity in diverse locations around the world and it was clear that all of these regions could not have had polar locations at the same time. He remembered Joseph Kirschvink's proposal of the freezing over of Earth from pole to pole. This provided a simpler explanation of the world-wide evidence of dropstone strata for the era of about 600 million years ago.

Hoffman quickly absorbed the limited amount of information Brian Harland and Joseph Kirschvink published on the notion of a total global freeze. What was needed was some irrefutable evidence for such a state of the world.

Hoffman and Shrag's first efforts in obtaining scientific proof looked into the carbon isotope ratio in the carbonate rocks that preceded and followed the strata of rock with evidence of glacial activity. Carbon in nature is a mixture of two isotopes; one with an atomic weight of twelve and the other with a weight of thirteen. These isotopes are exactly the same chemically and differ only slightly in physical properties that might affect such things as diffusion. Miraculously living organisms manage to absorb a higher proportion of lighter isotope than the heavier one. This leaves in the environment where the organisms feed a higher proportion of the heavier isotope. Therefore the ratio of C13 to C12 is an indicator of the presence or absence of living organisms. Hoffman and Shraq found no depletion of the C12 indicating that when the carbonate rocks were formed there were little or no living organisms feeding. This is plausible interpretation but opponents of the theory can come up with alternative processes to account for the evidence; such as massive flooding of the feeding environment of the organisms could have swamped the effect of the organisms feeding on the isotope ratio.

While Hoffman and Shrag were trying to develop evidence for the idea of Snowball Earth there were others trying to challenge it. One of the strongest bits of evidence confirming that the magnetic orientation data was the original orientation of the dropstone strata came from the Flinders Ranges in Australia. Drilled core samples of the glacial era rock collected with great effort by Linda Sohl of the Lamont Doherty Earth Observatory of Columbia University. She obtained a deep enough core that she could determine that the glacial activity had continued over such a long period that there had occurred several magnetic reversals. This meant that the rocks magnetic orientations were the original ones and the period of time involved was at least hundreds of thousands of years and probably millions of years.

In the relatively few years since Hoffman started promoting the Snowball Earth concept the preponderance of evidence has supported there having been a near-global glaciation. The quibbling has been over whether there might have been a band of open ocean near the equator (slushball Earth). Hoffman and Shrag tend to stick with the hard version of the theory but the acceptance of a soft version means they have won the intellectual dispute. The rest of the debate has been more involved with the refinement of the thesis rather than its being discredited. For example, the discovery of a geologic formation associated with freeze-thaw in the rock strata associated with the global glaciation did not disprove the thesis. Instead analysis revealed that under the conditions of Snowball Earth the seasonal variations would have been more extreme than they are now and land areas near the equator could have experienced summertime thaws.

However one must be cautious about presuming that it is only the land near the equator where thawing might occur. For land at the highest latitudes the sun is shining 24 hours a day during the summer. This results in those area having a higher input of solar energy than the lower latitudes lands. See Insolation.

Hoffman still sticks to the hard version of Snowball Earth although he is now in the minority among geologists. Most accept slushball Earth. A recent article in Science reports on studies of the Sturtian glaciation which occurred about 700 million years ago. That article, authored by Francis A. MacDonald and nine others including Daniel P. Schrag, concluded on the basis of study of rocks in what is now northern Canada that there was ice resting on ground below sea level when those rocks were in an equitorial position. Therefore, they conclude, the Sturtian glaciation was global in extent.

(To be continued.)


  • Paul F. Hoffman and Daniel P. Shrag, "Snowball Earth; ice entombed our planet hundreds of millions of years ago, and complex animals evolved in the greenhouse heat wave that followed," Scientific American, (January 2000), pp. 68-75.
  • "Considering a Neoproterozoic Snowball Earth," Science, (May 14, 1999), pp. 1087.
  • Gabrielle Walker, Snowball Earth: The Story of the Great Global Catastrophe that Spawned Life as We Know It, Crown Publisher, New York, 2003.
  • "The Earth Moves: The Cambrian Explosion," The Economist, (July 26, 1997), pp. 70-71.
  • Richard Monastersky, "When Earth tipped, life went wild (True Polar Wandering)," Science (July 26, 1997), pp.52.
  • Joseph Kirschvink, "Evidence for a large-scale reorganization of early Cambrian continental masses by inertial interchange: true polar wandering," Science, (July 25, 1997), pp. 541-545.
  • "Snowball or slushball?," Science, (November 1, 2002), pp. 925-926.
  • Naomi Lubick, "Snowball fights," Nature, (May 2, 2002), pp. 12-13.
  • Richard A. Kerr, "Snowball Earth has Melted Back to a Profound Wintry Mix," Science, vol. 327 (March 5, 2010), p. 1186.
  • Francis A. MacDonald et al., "Calibrating the Cryogenian," Science, vol. 327 (March 5, 2010), p. 1241-1243.

HOME PAGE OF applet-magic
HOME PAGE OF Thayer Watkins