Sunday, January 19, 2014

ExClimate: The Ice Albedo Feedback mechanism and Snowball earth

This article is a little bit of a tangent to my regular articles. I am presently doing an online course - Climate Change: Challenges and solutions - offered by the University of Exeter (UK). So please indulge me as I also use this blog for some climate course work. This article is for section 2.3 of the course on Snowball earth events that have ocurred in the distant past.

No one is sure what precipitated the onset of snowball earth events. The oldest snowball earth event occurred approximately 2200 million years ago. The most recent event ended about 635 million years ago and is thought to have lasted 6 to 12 million years. Very basic life managed to survive at least the most recent snowball earth event.

The earth's orbit around the sun, it's axial tilt, and wobble all have different millenial cycles which can affect the amount of solar radiation being absorbed by the earth. Earth’s orbit changes from near circular to oval shape on a 100,000-year cycle (eccentricity). Earth’s axis is tilted and is presently at 23.5 degrees. It varies between 21.5 and 24.5 degrees every 41,000 years (obliquity - Milankovitch cycles). As the Earth spins it wobbles on its axis toward and away from the Sun over the span of 19,000 to 23,000 years (precession).

What caused the earth to go into a giant snowball is still being debated in scientific circles. A number of theories have been advanced including :

  • a reduction in solar output,
  • the Earth passing through rare space clouds,
  • a combination of high obliquity in the Earth's orbit, orbital eccentricity or precession;
  • Lithospheric weathering reducing atmospheric carbon causing planetary cooling

There is some evidence for the last theory, that with less radiation from a dimmer sun combined with active rock weathering over millions of years sucking carbon out of the atmosphere into the oceans, resulted in reduced atmospheric carbon dioxide and methane levels. With less greenhouse gases to warm the earth, this started a climate cooling feedback process involving the ice albedo feedback effect.

Hoffman and Shrag (2002) argue that based upon the palaeomagnetic evidence of Joe Kirschvink, with the continents then located around the tropics and mid latitudes, this raised the planetary albedo. "Others had pointed out that silicate weathering would most likely be enhanced if many continents were in the tropics, resulting in lower atmospheric CO2 and a colder climate." they say in the abstract of their article.

Russian climatologist Mikhail Budyko formulated ground-breaking equations for the heat balance of the earth. His calculations revealed that once glaciation reached latitude 30 degrees, a runaway cooling effect would be in motion leading to freezing to the equator and snowball earth.

Geology Professor Paul F. Hoffman argued that glacial dropstones in sub tropical locations such as Namibia were the product of a Snowball Earth. But other scientists argued that continental drift could explain these signs of glaciation. The theory was that the continents had been located closer to the poles and subject to glaciation, and due to the theory of continental drift have since located across the globe including the tropics.

This puzzle was resolved by Professor Joseph Kirschvink at the California Institute of Technology through examining the minute magnetic fields in droprocks. Rocks have magnetic direction in their structure from when they are formed. Nearer to the poles, the magnetic direction is up and down while closer to the tropics it is more sideways. By examining each specimen through a powerful magnetometer it was possible to map the latitude where each rock was formed. This effectively dismissed the continental drift theory.

Paleontologist Guy Narbonne raised the problem of how could life survive, as it did in the fossil record, if ice ball earth covered the world's oceans with meters of solid ice. Prior to the last snowball earth the fossil record shows that microscopic and centimeter sized life forms existed. This included cyanobacteria such as stromatalites, and eukaryotes which includes alge and amoebae, and a few small worm like organisms.

The answer came through a variety of research from the deep sea and Antarctic environments.

Life thrives around volcanic vents in the Antarcticand the deep ocean where light cannot pierce. Such life would have survived any snowball earth event as it is not dependent on light from the surface.

Research in ice encased lakes in the dry valleys of Antarctica reveals there is plenty of light despite ice sometimes meters thick. One of the properties of slow freezing ice is that impurities like salt in salt water and dust are filtered out leaving ice that is highly transparent. This allows algae even under meters thick ice to keep photosynthesising.

Modelling by Dave Pollard and Jim Kasting published in 2005 - Snowball Earth: A thin-ice solution with flowing sea glaciers (abstract) - reveals that near the equator a zone about 2000 km wide may have existed where the ice depth was less than 2 meters thick. Around the tropics it may even have been an icy slush rather than thick sea ice. Such a zone would allow organisms reliant on photosynthesis to continue to survive in the oceans around the equator.

Volcanic emissions and snowball Earth

Once the feedback process starts and snowball earth results, how did the process reverse itself? Well, earth's plate tectonics and volcanism would continue and slowly CO2 emissions by volcanic activity undersea and by land based volcanos under ice sheets would build up slowly. The greenhouse effect would slowly come into play warming around the equators first, melting ice and slowly reversing the ice albedo as ice became open ocean.

The process is likely to happen with a great deal of force and 'cataclysm', with some climate modeling suggesting that the meltdown could occur in as little as 2000 years. A very rapid response in geological time. The resulting climate would be an ultra greenhouse transient before long term weathering and biological processes could again sequester the excess carbon in the hydrosphere and lithosphere.

Hoffman and Schrag's 2002 article in Terra Nova titled The snowball Earth hypothesis: testing the limits of global change describes this transient climate overshoot:

The intense greenhouse ensures a transient post-glacial regime of enhanced carbonate and silicate weathering, which should drive a flux of alkalinity that could quantitatively account for the world-wide occurrence of cap carbonates. The resulting high rates of carbonate sedimentation, coupled with the kinetic isotope effect of transferring the CO2 burden to the ocean, should drive down the Carbon 13 of seawater, as is observed.

The authors produce in Figure 7 an outline of a snowball earth event. Notice the abrupt onset and termination of glacial events, which are also associated with large shifts in carbon 13 in seawater, and the deposit of large carbonate layers (cap carbonates) with post glacial sea level rise.


So, what do I draw from Snowball earth? Firstly, the importance of the ice albedo feedback mechanism. This can work in either negative or positive modes. Sea ice in the Arctic is currently reducing at a rapid rate due to global warming. The ice albedo of the Arctic is rapidly reducing allowing greater radiative forcing to occurr, increasing warming, melting more ice in a vicious circle. The Arctic is warming at double the global average.

The second conclusion I would draw is the importance of tipping points. Once glaciation reaches 30 degrees latitude, it is like a switch is pushed and the earth heads for a new stability as a snowball. What rescues the earth from the snowball configuration is the earth's own internal heat, plate tectonics and volcanic activity which starts a long slow process of building up atmospheric greenhouses gases to reverse snowball earth. There are tipping points for a warming earth too.

We tend to think that natural processes are slow, but modelling suggests that the reversing of snowball earth conditions happened relatively quickly and with much violent weather. Climate scientists are still trying to come to grips with understanding the non-linear dynamics of ice sheet collapse. One of the reason that sea level projections are so broad is because we don't know whether it will be a linear or exponential rate of acceleration in ice loss from Greenland and Anarctic ice sheets. At 0.8C of global atmospheric warming we may have already be seeing the initial stages of the Greenland ice sheet collapse. The West Antarctic Ice sheet also is showing signs of instability.

Watch the BBC Horizon Snowball Earth (2001) Documentary (Course section 2.2). An interesting scientific detective story.


  • Paul F Hoffman and Daniel P. Schrag, Terra Nova (2002), The snowball Earth hypothesis: testing the limits of global change - Full Paper Retrieved 19 January 2014
  • David Pollard and James F. Kasting, Journal of Geophysical Research, VOL. 110, C07010, (2005) - Snowball Earth: A thin-ice solution with flowing sea glaciers, doi:10.1029/2004JC002525 Retrieved 19 January 2014
  • Dartmouth Undergraduate Journal of Science, In Spring 2010 / May 30, 2010 - Oceans of Ice: The Snowball Earth Theory of Global Glaciation -
  • Bill Maguire, Waking the Giant: How a changing climate triggers earthquakes, tsunamis, and volcanoes (2012), Oxford University Press
  • Paul F Hoffman, Snowball Earth website - Retrieved 19 January 2014

Answers to Questions from Recommended Reading, Course 2.5

What are climate change records?

Reference: David Parker, Climate Modelling Expert, Met Office (UK): Climate change and observations

Climate change records come from observational measurements, satellite measurements, and indirect measurements (proxies). Satellite measurements started in the 1970s to the 1990s and have provided an accurate basis for measurements of climate and atmospheric factors. Accurate observational measurement, for the most part go back 100 to 150 years. Measurements before then tend to be more haphazard and subject to biases and errors in data.

Marine data collection has been haphazard based on ship records. From 2000 a network of Argo floats was launched to collect data in the world's oceans from the surface to 2000 meters depth.

Scientists also use proxy measurements for data on changing climate from tree rings, ice cores and coral cores. While these aren't as precise as observation or satellite measurements, they can give an a good approximation of climate variables and how they have changed. See recent research that coral cores may help in extending the Pacific Decadel Oscillation cycles back in time up to perhaps 400 years.

Update 22 January: Just on climate change records and proxies... Found a strange one: red-to-green ratios from great painting masters of sunsets can provide independent proxy Aerosol Optical Depth (AODs) that correlate with widely accepted proxies and with independent measurements. Weird hey? But it appears to be true.

CF Zerefos et al 2013, Further evidence of important environmental information content in red-to-green ratios as depicted in paintings by great masters, Atmos. Chem. Phys. Discuss., 13, 33145-33176, 2013 doi:10.5194/acpd-13-33145-2013

How do volcanoes affect climate change?

Reference: NASA Earth Observatory - Aerosols: Tiny Particles, Big Impact

Volcanic activity adds aerosol particles (mainly sulfates) to the atmosphere and stratosphere as well as substantial quantities of carbon dioxide. The eruption of Mount Pinatubo in 1991 caused a drop in global average temperature for the following year of about 0.5C. Recent research in 2013 suggests moderate volcanic activity may be slowing global warming by 25 per cent.

Sulfate aerosols in the stratosphere have a net cooling impact over about a year or two reflecting some of the inbound solar radiation. There are various geo-engineering proposals to seed the lower stratosphere with sulfate particles which would help cool the earth, but such schemes may have unintended impacts on precipitation as well as not addressing ocean acidification issue. I discuss some of these unintended consequences in my recent article: Climate Geo-engineering study on sulphate injection shows Hydrological disruption to rain and severe drought

How is today's warming different from the past?

Reference: NASA Earth Observatory - How is Today's Warming Different from the Past?

Today's warming is happening at a much more rapid pace than previous periods of warming. This is evident from proxy records of paleoclimates. Over the last century the warming has been 10 times the rate as the most rapid warming during the ice age recovery. The Earth has already warmed by about 0.8C above pre-industrial times. It is projected to warm between 2 and 6C by the end of the century.

This has large implications for species adapting to changing climate envelopes. The velocity of climate change is such that many species will be driven to extinction, unable to move fast enough in latitude or elevation, or being blocked by human infrastruture or natural obstacles. Habitat loss and climate change causing a 6th mass extinction. A recent study from the University of Hawaii found that the world's ocean are already outside of natural variability, and most cities will pass their climate point of departure in the next 50 years. Ocean acidification is increasing at unprecedented rate not seen in last 300 million years, with marine scientists warning that Oceans at high risk of unprecedented Marine extinction.

What is the role of isotopes in determining temperatures from the past?

Reference: Gavin Schmidt, January 1999, NASA - Cold Climates, Warm Climates: How Can We Tell Past Temperatures?

Isotopes are important for providing measures of changes in the environment. Most people have heard of carbon dating which allows scientists to estimate the age of rocks and artifacts based upon the proportion of two carbon isotopes.

Similarly Oxygen isotopes of 18 O and 16 O are present in different ratios based upon the water temperature when the calcium carbonate shells of marine creatures are formed. This gives scientists an accurate way to determine water temperatures and to determine the origin in the tropics or mid- latitudes. The background ratio can tell us of global patterns of evaporation and precipitation, including the waxing and waning of major ice sheets.

While the isotopic ratio gives us information on temperature, evaporation, rainfall, with one measurement, it can be difficult discerning and disentangling the multiple effects.

How have trees been used to reconstruct different climate variables across the world?

Reference: Principles of Dendrochronology.

Tree rings provide a proxy measurement of past climate variables, including precipitation, temperature and proportion of cloud cover to identify the conditions in the growing season. Using cross dating in several related tree ring series allows for accurate identification of specification years. Site selection is also important to select trees on a site sensitive to the environmental variable you want to measure. Sample trees from the margin of the species Ecological amplitude (natural range and elevation) are often the most important to test. Multiple samples from the same tree and other trees in the vicinity can reduce the amount of incidental 'noise' in the tree ring record.

Trees which are growing in the Tropics all year are much harder to differentiate the growing cycles. Dendrochronology best works where there is a definite growing season. Trees from midlatitudes and high latitudes shut down growth during the winter season. While this makes reading the different concentric tree rings easy the tree rings really only tell you about the climate changes during the main growing period in spring and summer.

Many trees grow for hundreds and some for aup to a thousand years which provides proxy record over the recent past.

How can ice cores provide a record of atmospheric composition?

Reference: NASA Earth Observatory, (2005) Paleoclimatology: The Ice Core Record

Ice cores capture in their layers an annual record of temperature, precipitation, atmospheric composition, volcanic activity, and wind patterns. Thickness of layers can tell you the amount of snowfall, and comparing this to other cores nearby can also tell you wind direction and patterns. The chemical composition of each layer and measurement of the ratio of Oxygen isotopes can indicate indicates temperature details.