ABSTRACT

This is a collaborative proposal by Principal Investigators at the Harvard University and Hamilton College. They will study the glacial history of the two most recent "Snowball Earth" epochs, which preceded the first appearance of complex animal life on the planet. (Snowball Earth refers to a state in which the entire ocean is covered by Equatorward-flowing glaciers. It is the manifestation of a fundamental instability in the climate system due to ice-albedo feedback.)
Snowball episodes expanded millions of years because atmospheric pCO2 must reach hundreds of times present levels for deglaciation to occur. This is achieved because silicate weathering (which consumes CO2) is attenuated, but volcanic outgassing is not. Glacial deposits
(diamictite) previously used to test the snowball concept may actually have formed during
deglaciation, under very different conditions, creating a false dichotomy between theory and
observations. Fine-grained suspension deposits between or beneath diamictites were previously
regarded as interglacial or interstadial, but they may better represent marine sedimentation
during snowball Earth itself. When a snowball Earth deglaciates, ferrous iron dissolved in
seawater should precipitate as iron oxide. Iron-oxide-rich deposits are observed to follow
suspension deposits and precede diamictite, consistent with the snowball hypothesis if the former represent the snowball Earth and the latter its demise. The suspension deposits are locally
associated with carbonate sediments and other indications of open water. Models imply that
invasion by marine glaciers from higher latitudes will keep the tropical ocean covered by thick
ice even after the surface temperature reaches the melting point (due to greenhouse forcing).
Consequently, 'oases' will open up on tropical shelves and inland seas that are protected from
marine glacial invasion. Carbonates precipitated in snowball oases have characteristic isotopic
patterns, dictated by the large atmospheric carbon reservoir. These new concepts could eliminate
the most potent criticisms of the snowball Earth concept based on the glacial record. Stratigraphic mapping and logging will be done in the fjord region of East Greenland, northwestern Nordaustlandet (Svalbard), and the Mackenzie Mountains of the northern Canadian Cordillera. In these areas, the full glacial history of snowball Earth is well developed and exposed. Sample sets will be collected and curated for collaborative geochemical studies.

Intellectual merit. The proposal brings together two of the most experienced researchers in
their respective fields in an intensely collaborative field-based study to test a far-reaching theory
(snowball Earth) using creative and original concepts. Some of the concepts (e.g., snowball
oases) derive from recent theoretical modeling in yet another field (geophysical fluid dynamics).
A strength of the Harvard team's prior work on the snowball Earth problem was the strong
interest and involvement of a colleague, the geochemical oceanographer Daniel P. Schrag.

Broader impact. Both Principal Investigators hold appointments at liberal arts institutions, where the primary responsibility is to undergraduate instruction. Both incorporate elements of their research in courses they teach, and both include students as field and laboratory assistants. They regularly give lectures intended for the general public, and are frequently consulted by media science writers and television producers covering aspects of Antarctic geoscience and snowball Earth. They have also contributed to courses in science journalism. An article in Scientific American on Snowball Earth co-written by one of the Principal Investigators is used by many elementary school teachers in their classes.

Please report errors in award information by writing to: awardsearch@nsf.gov.