Wednesday, September 4, 2013

Snowball Earth: Hydrological Cycle

The Earth underwent at least two global glaciation events during the Neoproterozoic era - the Sturtian at ~720 Mya and the Marinoan at ~635 Mya (Pierrehumbert et al., 2011). Other estimates place the two events at ~740 and ~635 Mya respectively (Trindade and Macouin, 2007). These global glaciation events are more commonly known as Snowball Earth events where ice covered the Earth right to the Equator. During a Snowball Earth event, the thick global ice cover effectively eliminates the ocean’s thermal inertia. As a result, the low thermal inertia of the global ice cover resulted in large variations in surface temperature.

Figure 1: January surface air temperature for several general circulation model (GCM) simulations of a Snowball Earth with atmospheric CO2 at 2000 ppmv (parts per million by volume). Note that Earth’s pre-industrial atmospheric CO2 concentration is 280 ppmv, January corresponds to winter in the northern hemisphere and 273 K is equal to 0°C. (Pierrehumbert et al., 2011)

A sluggish hydrological cycle is expected on a Snowball Earth due to the low temperatures and ice covered ocean. The basic structure of a Snowball Earth hydrological cycle consists of a net ablation zone near the Equator where the annual mean precipitation minus evaporation (P-E) is negative. GCM simulations show that the existence of such a net ablation zone is robust. On a Snowball Earth, the ocean will be covered by a thick layer of ice, very much like a global version of Antarctica’s Ross Ice Shelf. The thick ice deforms under its own weight and flows as a sea glacier. Since it is colder towards the poles, the ice cover is thicker at the poles and thinner at the Equator. As a result, the ice will tend to flow from Pole to Equator.

Figure 2: Schematic of Antarctica’s Ross Ice Shelf.

Figure 3: Schematic of sea glacier flow in the weak (P - E) limit (lower panel) and strong P-E limit (upper panel). (Pierrehumbert et al., 2011)

There are 2 limits to the hydrological cycle on a Snowball Earth (Figure 3). The first limit is where atmospheric moisture transport is absent (weak P-E case) and the second limit is where atmospheric moisture transport is substantial (strong P-E case). In the weak P-E case, the sea glacier flows from Pole to Equator. This causes the ice at higher latitudes to be thinner than the local equilibrium value, so that new ice forms there by freezing onto the base. Consequently, the ice in the tropics becomes thicker than the local equilibrium value, so that excess ice melts at the base. Melt water from the tropics circulates through the ocean where it refreezes at the base of the ice at higher latitudes, thereby completing the hydrological cycle.

In the strong P-E case, ice freezes onto the base of the ice in the tropics and is brought upward to the surface, where it sublimates into the atmosphere and falls as snow at higher latitudes. The snowfall thickens the ice at higher latitudes to a thickness that exceeds the local equilibrium thickness. To compensate for the extra thickness, both basal melting of ice at higher latitudes and Pole to Equator flow of ice occur, which completes the hydrological cycle. In the strong P-E case, dust deposited together with snowfall at higher latitudes can be transported through the ice to the basal melting region. This delivers nutrient-laden surface dust into the ocean, allowing life to continue to thrive at localities that serve as refugia on a frozen planet (Hoffman and Schrag, 2000; Campbell et al., 2011).

- Pierrehumbert et al., “Climate of the Neoproterozoic”, Annual Review of Earth and Planetary Sciences 39 (2011) 417-60
- Trindade and Macouin, “Palaeolatitude of glacial deposits and palaeogeography of Neoproterozoic ice ages”, Comptes Rendus Geoscience 339 (2007) 200-211
- Hoffman and Schrag, “Snowball Earth”, Scientific American (January 2000), Volume 282, pp. 68-75
- Campbell et al., “Refugium for surface life on Snowball Earth in a nearly‐enclosed sea? A first simple model for sea‐glacier invasion”, Geophysical Research Letters Volume 38, Issue 19, October 2011