Allday Endowed Lecturer Hoffman shows Snowball Earth theory is no snowjob

April 5, 2006

AUSTIN, Texas—Paul Hoffman of Harvard University laid out evidence for one of modern geology’s great detective stories—the theory of global glaciation, or Snowball Earth—during one of his major presentations for the Edwin Allday Endowed Lecture Series at the Jackson School of Geosciences.

Speaking to an auditorium full of students and scientists March 30, Hoffman reviewed facts that “had long been known to exist” and recent scientific results that “renew the chance to resolve controversies” in the Snowball Earth theory.

First advanced by Joe Kirschvink of the California Institute of Technology in 1989, the Snowball Earth theory resolves a series of geologic anomalies—primary among them the presence of synchronous glacial deposits in equatorial regions—by positing that Earth experienced several complete freezes during the Proterozoic Eon (from 2.5 billion to 535 million years ago). The most recent two glaciations, according to the theory, occurred during the Neoproterozoic about 710 and 635 million years ago. Each episode is supposed to have lasted around 10 million years.

Such dramatic freezes conceivably would have wiped out all life on Earth. For this very reason, scientists initially thought the deep freezes could not possibly have taken place, since the fossil record shows that cyano-bacteria and microscopic organisms were living in the oceans before the most recent freezes.

Scientists postulate that cracks in dynamic sea ice created a refugium for phototrophs on Snowball Earth . See larger image. Photo courtesy of the Snowball Earth Web site.

New discoveries about the ability of aquatic organisms to sustain life under cracked ice packs, however, have eased doubts stemming from the fossil record. Meanwhile, since 1992, Hoffman and Harvard colleague Dan Schrag, a geochemist, have developed and disseminated a geological case that grows more and more convincing with each round of data and analysis.

Through the research of Hoffman, Schrag and others, Kirschvink’s theory has snowballed in credibility. An idea that once seemed absurd now stands on the verge of scientific orthodoxy.

Cold War Origins

Scientists today may be open to the idea of Snowball Earth, but the first researchers to contemplate the idea, a group of climate physicists in the 1960s, were confident that the scenario could happen in the future but had never happened in the past.

Climate physicists in Russia, Norway, and the United States first realized the theoretical possibility of global glaciation in the 1960s, said Hoffman. The unlikely impetus for their discovery was the Cold War. Scientists concerned about humankind’s ability to withstand a nuclear holocaust contemplated what would happen after a series of massive nuclear explosions. Lingering soot and smoke clouds, they theorized, would block the sun, dramatically cooling Earth.

Climate physicists quickly realized that the nuclear winter would not necessarily stop with an initial cycle of cooling. Ice’s albedo effect—its ability to reflect back the sun’s heat—could cause the planet to freeze completely, and with remarkable speed.

Slide illustrating ice-albedo, a critical factor in snowball models . See larger figure. Figures courtesy of the Snowball Earth Web site.

Albedo is a measure of reflectivity—the fraction of the sun’s radiation that a surface or body reflects back. Bare ice has a much higher albedo (~0.45-0.65) than open water (~0.1) or bare land (~0.3). Bright white snow has an even higher albedo (~0.9). Ice and snow, therefore, reduce the sun’s warming properties in comparison with bare land and open water, which absorb and retain more of the sun’s heat.

The cycle of Earth’s climate as we know it today checks the ice-albedo effects of the polar caps, thanks in large measure to greenhouse forcing, the cycle of warming caused by carbon dioxide and methane in the Earth’s atmosphere. But this would not be the case if a critical mass of ice were to form. While running basic climate scenarios, the Cold War-era scientists realized that if just half of the Earth crusted over in ice, the increased cooling would throw the climate equilibrium out of whack. Not enough greenhouse gases would return to the atmosphere, which in turn would lead to runaway ice-albedo feedback: cooler temperatures would beget more and more ice.

“When you get to the point that Earth is half covered in ice,” said Hoffman, “you get unstoppable feedback. Within one to two centuries, the ice lines meet at the equator.”

Impossible dream or plausible nightmare?

Climate physicists in the 1960s were confident “that such a freeze could happen, but hadn’t happened,” said Hoffman, because if it had, the freeze would never have been reversed. They believed that in a completely frozen Earth, greenhouse gases could not build up enough to force the cycle of warming and eventual melting that, say, led to the thawing of the Pleistocene Ice Age. According to this view, had Earth ever become a complete snowball (something which did not occur during the Pleistocene), it would still be one today, and all life would have long ago become extinct.

Kirschvink’s first major contribution to the Snowball Earth theory was to show that the prevailing models of global glaciation were incorrect in their predictions. Far from being irreversible, a Snowball Earth would be self-reversing, albeit on a very slow clock. Even under frozen conditions, Kirschvink pointed out, plate tectonics continue. Volcanic eruptions spew forth greenhouse gases. At the same time, snow, the dominant precipitation on a frozen planet, does not remove carbon dioxide from the atmosphere at anything near the rate that rain does under temperate conditions. As a result, carbon dioxide would build-up over millions of years, eventually warming Earth enough to cause melting.

And that melting would be dramatic. A reverse of the ice-albedo effect could erase a ten-million-year-old coat of ice in just a few centuries, said Hoffman—centuries of raging winds, waves, and storms dwarfing the tsunamis and category-five hurricanes of the present day.

Present global distribution of older (“Sturtian”) Cryogenian (ca 720-700 Ma) and younger (“Marinoan”) glacial and glacial-marine deposits (ca 660-635 Ma). Red dots indicate deposits that include banded iron or iron-manganese formations (BIF). See larger images. Figures courtesy of the Snowball Earth Web site.

Kirschvink not only suggested this “freeze-fry” cycle could happen—it had happened at least three times, based on the geological record.

Building the case

During his talk, Hoffman described the three snowball episodes in the geologic record, which took place circa 2220 Ma (“Makganyene”), circa 710 Ma (“Sturtian”) and circa 640 Ma (“Marinoan”). (A fourth major but regional episode known as the Gaskiers glaciation occurred about 580 million years ago in eastern Newfoundland in Canada.)

He then reviewed highlights of the principal evidence for the episodes. The “classical arguments,” he said, included:

• Carbon isotopic data showing large changes in the carbon content of sea water bracketing the snowball/glaciation periods.
   
• Paleomagnetic data indicating globally distributed glacial deposits in mid and low latitudes synchronous with the glaciation periods.
   
• Tidal bundles implying a shallow marine depositional environment and the formation of an iceline a few degrees from the equator.
   
• Banded iron formations found exclusively in glacial marine strata after 1.9 Ga, indicating widespread anoxia, which is consistent with an ice-covered ocean, and an increase in the ratio of iron to sulfur entering the ocean, consistent with ice-covered continents.
   
• The deposition of post-glacial “cap carbonates” with unusual sedimentary structures on continental margins and inland seas globally in the aftermaths of the Sturtian and Marinoan glaciations. Their origin remains controversial but they are predicted to occur in the greenhouse aftermath of a snowball earth.

(The Snowball Earth Web site contains a more detailed explanation of the evidence in support of the theory. See, “What is the evidence for snowball earths?”)

Hoffman also touched on some of the new results confirming the theory. Prominent among these are recent data suggesting changes in seawater pH relative to the freeze periods and evidence of extraterrestrial dust trapped in ice for millions of years and released upon deglaciation.

Today, most scientists agree, said Hoffman, that the periods identified as global glaciations were extreme states. Controversies remain, however, such as the extent of ice that formed in the tropical oceans. The scientific literature presents three models: Complete glaciation thick enough to rule out photosynthesis beneath the ice; complete but less severe glaciation, leaving a 2000 km equatorial zone where ice was less than 2 meters thick; and nearly complete glaciation, leaving a thin equatorial sea between the ice lines.

Concluding his talk, Hoffman addressed one of the most fascinating aspects of the theory: the potential impact that the glaciations had on biological evolution. The topic is particularly interesting given that eukaryotic cells first appeared after the penultimate Snowball Earth and multicellular life forms began to appear not long after the last glaciation.

Figure showing early animal diversification after the Marinoan glaciation . See larger figure. Figures courtesy of the Snowball Earth Web site.

Did the freeze-fry cycle contribute to the evolution of life on Earth? Was the end of the last freeze the thaw that opened the way for complex organisms?

The fossil record provides no support for close correlations in time between the glaciations and the evolution of organic life, said Hoffman, and causal links are entirely speculative. But audience members could hear the enthusiasm in Hoffman’s voice as he shared the possibility that the second-to-last glaciation played a role in the evolution of eukaryotes, while the Marinoan glaciation contributed to the achievement of multicellularity in microscopic animals.

The increasing rate of speciation following the last snowball episode offers another tantalizing clue of the freeze’s contribution to organic life. Showing a chart on early animal diversification, Hoffman noted that the explosion of life forms reflected in “the phyletic tree suggests the chain of events leading to multicellular animals might” have had roots in the Neoproterozoic snowball episodes.

Returning to firmer ground, Hoffman stated that while the explosion of complex life forms can not be tied with certainty to Snowball Earth, an explosion of papers definitely can be tied to the appearance of a paper on the topic in Science in 1998 (“A Neoproterozoic Snowball Earth,” P. F. Hoffman, A. J. Kaufman, G. P. Halverson and D. P. Schrag, Science, Vol. 281, pages 1342-1346, 28 August 1998). The paper helped lead a once controversial subject into the mainstream, where Snowball Earth is now a respectable topic for scholarly discussion.

“Some people think Snowball Earth isn’t that exciting anymore because people aren’t shouting and screaming,” said Hoffman. “But now is the time it’s most exciting because we’re seeing real science.”

For more information about the Jackson School contact J.B. Bird at jbird@jsg.utexas.edu, 512-232-9623.