What Does it Take to Ice the Earth?
Using a global climate model, explore the range of system boundary conditions that would produce a Neoproterozoic "snowball Earth."
Sarah Das and Mark Behn
Water-filled fracture propagation in ice.
A significant question in ice sheet dynamics is how and under what conditions the surface meltwater can penetrate to the bed of the ice sheet causing a dynamic response. Recent studies suggest that climate variability can induce rapid changes in ice sheet behavior. These timescales (1-10 years) are much too short for thermal diffusion to be important and suggest that water is rapidly supplied to the bed via water-filled cracks. This project would combine observation and theoretical constraints to better understand the initiation, propagation, and duration of fluid-filled cracks in subfreezing ice.
Jerry McManus and Delia Oppo
Drifting ice is geologically ephemeral, yet icebergs may play important
roles as climatic indicators and even as agents of climate change
through their influence on ocean circulation. Projects are available in
two very different geographic and climatic settings:
- The North Atlantic. During the last ice age, catastrophic iceberg discharges from North America choked much of the Atlantic, drifting south to the subtropics and east to the Iberian margin, and leaving a trail of characteristic debris on the seafloor. Icebergs also advanced episodically from other locations, including Iceland and the British Isles, with distinctive isotopic and petrologic signatures. The relative timing and magnitude of these discharges provide clues that may resolve whether the events were driven by regionally coherent climate change or by ice sheet dynamics influenced by the disparate basal conditions of the respective ice caps. Abundant icebergs were generally associated with the coldest conditions within the ice age, and it is hypothesized that the freshwater from the melting bergs acted to diminish the ocean’s climatically important overturning circulation, thus providing a positive cooling feedback. Student projects may focus on geochemical, radiochemical, petrologic and/or sedimentological aspects of iceberg discharges and their oceanographic and climatic impacts.
- The South Pacific. Prior to the last ice age, the world was generally slightly warmer than today, with sufficiently less global ice so that sea level was several meters higher. The Earth then descended into one of the most extreme glacial episodes of the last quarter billion years. This is the last time an interglacial interval such as the one we live in gave way to a new ice age, and although the transition to glaciation is widely believed to have been paced by insolation, it was not monotonic or globally synchronous. Several lines of evidence suggest that cooling occurred rapidly, and that the Southern hemisphere led the way, in contrast with the North Atlantic, which remained warm as ice sheets began to grow. The Andean glaciers of western South America are sensitive indicators of climate change, and debris from the icebergs that result when the glaciers reach the sea has the potential to serve as an indicator of glacial inception. Using sediments from the Chilean margin, a comparison of ice-rafted debris and the planktonic oxygen isotope indicator of meltwater can be made in the same core with the benthic oxygen isotope record of global glaciation to determine whether cooling over South America preceded that of the North Atlantic.
Rifting on Ganymede.
Bright terrain on Ganymede, an icy satellite of Jupiter, displays regularly spaced faults and longer-wavelength topographic undulations. These length scales of deformation can be used to constrain the thermal structure of the icy shell. In this project, you will use existing Finite Element codes to produce synthetic fault patterns and relate rift morphology with ice properties.
Glacier seismic cycle.
Shridar Anandakrishnan showed to us how the different dynamics of glacier b and d in Antarctica may relate to the frictional properties of the basal interface. We will build more realistic 2-layer spring-slider models for the b glacier considering a velocity-weakening surface (the frozen bed) over either a velocity strengthening or viscous surface (the till).
Wax experiments and lava tubes.
We have developed a laboratory experiment using liquid wax as a model of a lava tube.
The idea is to learn to estimate how far lava can travel through solid material within a
tube before freezing. There are also drainage tubes of water within glaciers with similar
dynamics. Existing theories of lava tubes usually involve a pre-selected radius for the
tube. The calculations do a good job finding the flow profiles and pressure drop for
various lava rheologies. However, the radius of an actual lava tube is a free parameter that is somehow selected in a balance between melting, flow rate, temperature and other
parameters. Moreover, the distance of travel is in some sense a free parameter too. I
want to understand their selection processes. Thus, we have made an experiment where
the size of a “tube” of melted wax is a result of the experiment instead of an imposed
A disk of aluminum at a fixed temperature was carefully leveled so that its central axis was vertical. Above this disk was a thin air gap and above the gap was a polycarbonate lid. An experimental wax (1-hexadecene) was injected at steady rate through a central hole with the aluminum below the freezing point. The wax spread out and as time progresses formed a sequence of frozen fans in the gap. After the gap became filled with these fans, the wax forced the lid upward and flowed out under the lid as a uniformly diverging radial sheet flow. Suddenly, a drainage channel formed in the ambient wax extending from the central hole to the outside rim of the cylinder. All of the flow became accommodated by the channel and the thin sheet-flow layer froze. Then, the flow became steady and drainage could continue indefinitely with this steady flow. We don’t have a very good theory predicting the width of this channel yet, but the size of the channel gets smaller as pumping rate is reduced. Caleb Mills performed measurements of the size last year. Now, we would like to see two things. First, more clearly whether there is a minimum flow rate, which would result in a frozen channel if reduced. No minimum is reported yet, but one would think that a very sufficiently small flow would not supply enough heat to counteract conductive cooling. Second, we have a hollow cylinder with cooled walls. Wax inside this would be more like a tube and we would like to try and form one.