Thursday, September 5, 2013

Snowball Earth: Thawing a Snowball

During the Neoproterozoic era (~1000 to ~540 Mya), the Earth experienced at least two global-scale glaciations at ~740 and ~635 Mya, where glaciers covered most of the Earth’s surface down to the deep tropics (Trindade and Macouin, 2007). Each global-scale glaciation or Snowball Earth event lasted for several million years. Over those millions of years, atmospheric CO2, a greenhouse gas, is expected to accumulate to immense levels due to continuous emission by volcanic activity and greatly reduced weathering on a frozen planet.

Figure 1: CO2 emitted by volcanos accumulates in the atmosphere of a Snowball Earth. The removal of CO2 from the atmosphere is limited by the absence of rainfall. Furthermore, the drawdown of atmospheric CO2 is almost nonexistent because silicate weathering is greatly reduced due to the overlying ice cover and very cold ground temperatures.

It is believed that the Earth thawed from a snowball state when the amount of atmospheric CO2 reached a level high enough to generate a sufficiently strong greenhouse effect. Evidence show that the amount of atmospheric CO2 towards the end of a Snowball Earth event accumulated to ~10 percent concentration; equivalent to ∼0.1 bar partial pressure (Bao et al., 2008). However, most climate models show that ~0.3 bar partial pressure of atmospheric CO2 is required to deglaciate a Snowball Earth with thick tropical ice cover (Pierrehumbert, 2004, 2005; Le Hir et al., 2007). This indicates that the accumulated amount of atmospheric CO2 during a Neoproterozoic snowball event was insufficient to cause deglaciation of a Snowball Earth. Such a discrepancy suggests that another mechanism in addition to the build up of atmospheric CO2 also played an important role in the deglaciation process.

In a study by Abbot and Pierrehumbert (2010), the authors show that the unique climatic conditions during a Snowball Earth event will allow a dust layer to develop over the ice surface in the tropics. This will cause the ice surface to be less reflective, allowing for greater absorption of solar radiation in the tropics and enabling deglaciation to occur with less than ∼100 mbar partial pressure of atmospheric CO2.

The three main sources of dust that can lead to significant deposits on the ice surface during a Snowball Earth event are: continental dust, volcanic dust and cosmic dust. Large non-glaciated continental areas are likely to exist on a Snowball Earth due to a very weak hydrological cycle. Such areas are probably more prevalent near the tropics where there is net ablation (annual average evaporation greater than precipitation). Additionally, palaeogeographic reconstructions indicate that much of the Earth’s continental masses after the break-up of the supercontinent Rodinia at around 800 to 600 Mya happened to be grouped along the tropics. Any non-glaciated continental area on a Snowball Earth is expected to produce a large amount of dust since it will be extremely dry, devoid of vegetation and subjected to huge diurnal temperature cycles that can cause soil fracturing and cryogenic weathering. Accumulation of continental dust is estimated to be on the order of 1 to 10 m/Myr.

For volcanic dust, the estimated rate of accumulation is ~1 m/Myr. Volcanic dust is important because it represents the minimum dust flux even if the continents are all glaciated and not contributing dust. The contribution from cosmic dust is negligible because its accumulation rate is estimated to be only ~0.1 mm/Myr. Given these estimates, the globally averaged dust accumulation rate during a Snowball Earth is reasonably estimated to be around 1 to 10 m per million years.

Simulations performed in this study show that the low latitudes on a Snowball Earth is a zone of net ablation where dust blown off the continents and from volcanic emission will accumulate over the ice surface, forming a tropical dust strip around the planet that is meters thick. Such a tropical dust strip makes the tropics less reflective to incoming solar radiation. As a result, temperatures will be somewhat higher than for a purely ice-covered surface, allowing the Earth to deglaciate from a snowball state with no more than ~0.1 bar partial pressure of atmospheric CO2.

Figure 2: Simulations of a Snowball Earth with surface air temperature in °C for January (left) and for the annual mean (right) as a function of longitude (horizontal axis) and latitude (vertical axis, which is linear and stretches from the South Pole to the North Pole). Here, the partial pressure of atmospheric CO2 is 0.0001 bar (a and b) and 0.1 bar (c and d) (e and f). The continental outline is represented as a thick black line. To simulate a tropical dust strip (e and f), the continental region is extended to include all the area within 15° of the Equator. Deglaciation, starting in the tropics, is likely to occur when the annual mean surface air temperature warms to roughly -10°C or more.

- Trindade and Macouin, “Palaeolatitude of glacial deposits and palaeogeography of Neoproterozoic ice ages”, Comptes Rendus Geoscience 339 (2007) 200-211
- Bao et al. (2008), “Triple oxygen isotope evidence for elevated CO2 levels after a Neoproterozoic glaciation”, Nature, 453 (7194), 504-506
- Pierrehumbert (2004), “High levels of atmospheric carbon dioxide necessary for the termination of global glaciation, Nature, 429 (6992), 646-649
- Pierrehumbert (2005), “Climate dynamics of a hard Snowball Earth”, Journal of Geophysical Research, 110, D01111, doi:10.1029/2004JD005162
- Le Hir et al. (2007), “Investigating plausible mechanisms to trigger a deglaciation from a hard Snowball Earth”, Comptes Rendus Geoscience 339 (2007) 274-287
- Abbot and Pierrehumbert (2010), “Mudball: Surface dust and Snowball Earth deglaciation”, Journal of Geophysical Research, Vol. 115, D03104. doi:10.1029/2009JD012007