The geoblogosphere spawns semi-monthly collections of blog posts on a particular theme, and this time around, Dr. Lemming is hosting with the theme of "things that make you go Hmmmm." The idea here is to write a blog post about something you don't understand in geology -- a mystery. Here's my contribution:

When I was in graduate school at the University of Maryland, I started hearing about a crazy notion that the entire planet had frozen over in the past. Apparently, multiple streams of evidence (chemical, isotopic, geologic, and magnetic) suggested that during the Neoproterozoic era of geologic time, the planet experienced a mega-Ice Age. There were even glacial deposits within a few degrees from the equator. If you've got glaciers operating within a few degrees of the equator, some scientists argued, then that means the Earth would have been entirely sheathed in ice. Its reflectivity ("albedo") would have been so high that most (~85%?) of incoming solar radiation would have been reflected back out into space, and that would have made the planet even colder, promoting more snow and ice. This positive feedback cycle would have reached a tipping point if the planet were covered in ice from the poles to approximately 30 degrees latitude: once it got that white, the "runaway albedo" feedback would have reached a tipping point, and wham, you've got a planet that looks like a great big snowball.

This led Joe Kirschvink (of Cal Tech) to dub this episode of glaciation the "Snowball Earth," which is about as catchy a name as a scientific hypothesis is every likely to get. The idea was then heavily promoted by Paul Hoffman (of Harvard), who was seeing strange stratigraphic patterns during field work in Namibia. Among the evidence Hoffman eventually accumulated for the Snowball were the following: "dropstones" (boulders, presumably dropped by icebergs into fine-grained offshore marine deposits, squishing the layers beneath them); conformable stratigraphy of "tropical" carbonate topped by glacial tillites, topped by more "tropical" carbonate; carbon isotope anomalies in overlying "cap" carbonates indicating a massive inorganic dumping of precipitated CaCO3; delicate crystal fans (some meters tall) precipitated rapidly in the post-Snowball ocean; and the temporary reappearance of banded iron formations (BIFs), which had not been seen since the Paleoproterozoic (and indicated an anoxic ocean, such as one sealed beneath a layer of ice).

When Kirshvink pitched the initial hypothesis, he also proposed how the Snowball could have ended (in a deliciously short, non-peer-reviewed paper): he noted that just because the surface of the planet was frozen, that would have meant diddly to plate tectonics. Radiogenic heat from the Earth's interior would have continued to drive plate tectonic processes, and that meant subduction would have continued, beneath the icy rime. If subduction continued, that meant that volcanoes would have continued to erupt, and as Iceland and Antarctica show us today, volcanoes can erupt underneath glaciers. This is important because volcanic outgassing has a substantial percentage (~15%) of carbon dioxide (CO2), and CO2 absorbs reflected infrared radiation: it's a greenhouse gas.

But with the entire surface of the planet frozen, what would have happened to this degassed CO2? If the planet's surface is frozen solid, that means the hydrologic cycle would be shut down, and the usual means of removing CO2 from the atmosphere (e.g. photosynthesis & also deposition of carbonate sediments like limestones) would be non-functional. Any CO2 emitted by volcanoes would therefore likely linger in the atmosphere, building up in concentration over time. Eventually, Kirshvink suggested, it built up to levels that caused global warming which compensated for the ice albedo effect, and the absorption of all that radiation by the CO2 melted the Snowball.

As evidence for this audacious idea, Kirshvink pointed to the cap carbonates: all that limestone ("cap carbonate") deposited on top of the glacial units needed a lot of CO2 to be dissolved in seawater (and a lot of Ca+ too). The cap carbonates, it was suggested, represented the stratigraphic removal of all that built-up CO2 from the atmosphere. Once the levels of CO2 were drawn down to a non-hothouse level, the cycle could repeat itself. Modeling calculations suggest that it would take about 5 million years of CO2 buildup to melt the Snowball.

And this is what I don't get: if you've got an atmosphere full of CO2, I can see how that would melt the Snowball. But wouldn't it then acidify the ocean (with carbonic acid, like we're seeing today), making calcite dissolve, rather than be precipitated? If the ocean is undersaturated with respect to CaCO3, then that ocean should not host accumulations of limestone. How could the voluminous worldwide cap carbonates be deposited in an acidic ocean?

On the Snowball Earth website, a list of suggested reasons why Snowball Earth could not have happened are listed, along with Hoffman, et al.'s scientific rebuttals. But when they come to the question of acid oceans and the deposition of cap carbonates, you can almost see them shrug: "These are serious criticisms," they note. Hmmmmm.

Post-script: The idea is intriguing not merely scientifically, but also in terms of the way science gets done: by people, sometimes people with outsized personalities. Paul Hoffman promoted the idea with an "evangelical zeal" (according to Gabrielle Walker, who wrote a book about the whole idea and the scientists involved). Hoffman's relentless pushing of the idea ruffled a good many feathers. Some scientists fought back, motivated in part by these chafing interpersonal dynamics. There's nothing like a little scientific controversy, and this is what Walker's book focuses on, more than the details of Snowball science.

When I found that Jay Kaufman (of UMD-College Park) was interpreting a local diamictite(near Aldie, VA) as a Snowball Earth tillite (and the overlying marble layer as a cap carbonate), I thought "this could make a great class." Last spring, I applied for and received a grant from the Virginia Community College System to develop a 2-credit class for NOVA utilizing these local rocks as a gateway to understanding the Snowball Earth hypothesis. I offered the class for the first time last summer, and I'll be offering it again this summer in August. We were fortunate to get rock samples from Virginia's two putative Snowball deposits as well as a suite of samples on loan from Gene Domack of Hamilton College. These "Snowball Suite" samples include tillites and dropstones from Namibia, Greenland, Mauritania, and Canada, as well as international BIFs and cap carbonate samples. I have to tip my hat to Dr. Domack and his colleagues: making these samples available is a terrific service in support of geoscience education.

Labels: climate change, CO2, geology, global warming, nova, snow, snowball earth, teaching