The latest ice age started a little over 2.5million years ago and consisted of numerous glaciated periods (glacials) in which continental glaciers covered greater or lesser parts of Eurasia and North America. These were interspersed with interglacial periods (interglacials). Early in the present 2.5 million year glaciated age (Quaternary), the severity of the icy periods was relatively mild (compared to recent periods) and the cycle lasted about 41,000 years. Around a million years ago, ice periods began to be more severe and to last around 100,000 years. The end of the glaciated periods appears to be synchronous with one of the Milankovitch cycles, namely the variation in the tilt of the earth (Obliquity) which has a 41,000 year cycle. The recent, longer glacials we have had for the past million years are still synchronous with the Milankovitch cycle but only every third or so Milankovitch nudge manages to kick the world into an interglacial.
Some exquisite work done by Larry Edwards et. al. of the U of Minnesota using corals and stalactites# has dated the ends of the ice periods much more accurately than previously and allowed a much clearer picture of when the glacials and interglacials within the latest ice age occurred. It is now clear that Carbon dioxide concentration in the air rises steeply as each glacial ends and an interglacial starts. Carbon dioxide is then sequestered, slowly declines in the atmosphere and new continental glaciers begin to grow.
#See New Scientist 22May 2010 p32
Despite the great improvement in dating, it still isn't clear if the sharp rise in Carbon dioxide precedes the end of an ice age or is a result of it. It seems unlikely, though, that some source of Carbon dioxide suddenly increases, triggering the beginning of an interglacial exactly in sinc with the Milankovitch obliquity. There are, however, a number of feasible scenarios that could explain the rise in Carbon dioxide as a result of the melting. Dating is not precise enough yet to definitely establish which came first This blog explores some of the possible mechanisms by which melting ice could give rise to massive increases in CO2
Another question is why the ice started melting with the second or third nudge from the Milankovich cycle but wasn't triggered by a couple of previous ones. One theory is that as the ice accumulates, it pushes down the land and hence the top of the glacier is at a lower altitude. Since the basalt basement on which the continents float has a specific gravity of about 3, when you add a kilometer of ice with a specific gravity of about 1 on top, it will sink a third of a km down. Put around the other way, every km of ice you add raises the top of the ice by 2/3 of a km. The sinking explanation seems a tad unlikely. The sinking takes time and even now, 11,000 years after the end of the most recent period of continental ice, land is still rebounding. One tends to think that there has to be more to what triggered the end of an ice period and some ideas will be presented in a future blog. This blog is concerned with which mechanisms could have led to a spike in Carbon dioxide once the melting had started. Such sources of Carbon dioxide would keep us in an interglacial period until various sinks had time to remove sufficient carbon from the air to allow snow to once more accumulate.
This blog is speculation. Like any hypothesis, one looks for tests to apply to see if they support or weaken the argument. For instance, one of the predictions of Einstein was that light from a distant star would be bent as it passed by a heavy object like the sun. This was tested during an eclipse of the sun. Stars that were made visible very close to the sun were seen to change their apparent position. The change was consistent with Einstein's predictions. While this was not unequivocal proof of Einsteins theory of gravitation, it did strengthen it. In this blog, after suggesting a result, I will suggest a "test" or "observation" which would strengthen or weaken the hypothesis.
Below are some possible sources of carbon dioxide caused by the melting of the Continental ice sheets.
It is well known that volcanoes release large quantities of Carbon dioxide. The source of at least some of this carbon dioxide, is the calcium carbonate that is heated when one tectonic plate sub-ducts under another, carrying with it the calcium carbonate that has accumulated on it. The accumulation of Calcium carbonate on the bottom of the ocean is one of the sinks for Carbon dioxide and volcanoes recycle this carbon back into the atmosphere.
The volcanism which has been experienced since man has recorded such things has been relatively mild. However there is ample evidence for giant caldera forming volcanoes such as Yellowstone in America, Lake Toba in Indonesia and Lake Tapo in New Zealand. There is also evidence of massive volcanism of the sort that created the traps in India, South Africa and South America. There is, however no evidence that any of these giant events occurred in sinc with the end of the numerous glacials over the present ice age.
However the following could have happened. Many magma's contain a lot of dissolved CO2. Just like in a bottle of soda, pressure keeps this gas in solution. Release the pressure and the gas begins to come out of solution. With three km of ice lying on top of the land, the added pressure would have an equivalent weight to about 1.2km of continental rock (sg 2.5). Removing this weight would have been like the rock slide on Mt St Helena which removed the pressure on the underlying magma and allowed a massive outpouring of volcanism. A bit like taking the lid of a pressure cooker immediately after removing it from the stove. (don't try this at home).
Ancient volcanism is visible in core samples from the bottom of lakes and oceans and sulphate from volcanism is captured in ice cores. If there was increased volcanism at the end of the last glacial or even at the beginning of the Eemian interglacial, their signature might be visible in such cores. In addition, the carbon that comes out of volcanoes will be old carbon. That is to say, carbon that is poor in C14. If carboniferous samples are available with independent dating from the beginning of the present interglacial, a C14 anomaly might be seen for he end of the recent glaciation*.
*Note: at a push, Carbon dating can date back 50,000 years. The end of the recent glaciation is within this range (11,000 years) but previous ones are not.
Suppression of Phytoplankton
Variation in Carbon dioxide in the atmosphere over a year is about 7ppm. At present we have a yearly cycle of 8ppm up and 6ppm down as we add fossil carbon dioxide. Imagine if each rise was not followed by a fall.
Phytoplankton growth depends on sunshine and a supply of nutrients. If it has both, phytoplankton grows at phenomenal rates. Phytoplankton take up Carbon dioxide to build it's substance.To get an idea of the magnitude of this effect, consider the productivity of Anchovy in the waters off Peru in La Nina years when the upwelling of nutrient rich water is in full flow. This fisheries provides much of the fish meal for the livestock trade of the world. When you consider that Anchovy are at the third tropic level (they eat zoo plankton which eat phytoplankton) and that only 10% of the mass from one tropic level is captured in the next level, it is clear that the production of algae is 100 times the production of Anchovy. At this site, it is noted that the rate of increase in carbon dioxide today depends on the El Nino - La Nina cycle. We are putting masses of Carbon dioxide into the atmosphere and as shown by the analysis from ManaLoa, it increases spasmodically, averaging about 2ppm per year. When there is upwelling off the coast of Peru, carbon dioxide increase is less than when the upwelling is not occurring. This is a small area when compared with, for instance, the whole Atlantic ocean. Imagine the effect of greatly reducing carbon dioxide uptake by phytoplankton over much of the ocean. Here, instead of invoking a source of Carbon dioxide as the ice begins to melt, we have the suppression of a sink. Same effect. So how would it occur.
At present, much water is evaporated in the warm climate around the Gulf of Mexico but the resulting saltier water is warm enough not to sink. It flows on the surface northward in what is called the Gulf Stream. As it travels north, it cools and eventually is heavy enough to sink. Added to this is the effect of the freezing of sea water in the North Atlantic and Arctic ocean. Freezing crystallizes fresh water ice from the sea water leaving behind cold saltier water. This also powers the sinking of cold surface water. The cold salty water from both of these sources flows south along the bottom of the ocean. The flow rate of the Gulf Stream is estimated at about 30million cubic meters per second so the return flow will be if a similar magnitude.
As masses of ice begin to melt the resulting fresh water flows into the ocean and floats on top. This, it is believed, would shut down this system of sinking water and stop the Gulf stream. The flip side of sinking water is that water has to rise somewhere. When the Gulf stream is operating, the heavy water flowing southward along the bottom of the ocean picks up nutrients from the rain of organic material from the surface. Somewhere in the oceanic circulation system, this water surfaces. The primary productivity powered by this system must be enormous and in fact, far greater than the Peru upwelling. Shutting it down would eliminate this primary productivity and hence its absorption of CO2.
In addition to fresh water pouring into the north Atlantic from the St Lawrence and other coastal rivers, the Mississippi system would transfer masses of water into the Gulf of Mexico. It could well be that the melting of the continental glaciers would, to a large extent, stop the overturn of the oceans. A similar situation would occur around Eurasia with the melting of her continental glacier.
Signatures of this may be present in ocean bottom cores. One might find a great reduction in fish scales in mud cores from where the water used to return to the surface. There also might be reduced (chemically speaking) layers of mud if the bottom of the ocean became anaerobic due to the lack of circulation.
Release of Clathrates
Clatrates are curious substances. They form when water and certain gases are mixed under pressure. Here we are concerned with methane clathrates. When methane is mixed with water, under pressure, it forms an ice. With sufficient pressure (4000m of sea water), a methane clathrate (methane hydrate) can form at up to 30 degrees centigrade. The higher the pressure, the higher the temperature at which a clathrate can form. The minimum pressure needed is the equivalent of about 300m of water and at this pressure, methane clathrate will form at a couple of degrees above freezing.
Of importance for our argument is that once a few hundred meters of ice have accumulated, the conditions are created at the bottom of the ice for the creation of clathrates. A clathrate contains considerable amounts of methane. A liter of methane clathrate, for instance, can contain as much as 160l of methane (measured at STP).
The question then becomes, are there sources of methane that would accumulate as clathrates under a forming ice cap, once the ice thickness had reached a few hundred meters deep. If there are, all this carbon would be released when the ice sheets melts. This would put the powerful green house gas, methane, into the atmosphere. The half life of methane is about 7 years. It combines with the oxygen of the air and forms Carbon dioxide. On a geological time scale, the methane is instantly converted to carbon dioxide and would appear as such in ice cores. However, during this transition period, green house warming could be strongly accelerated by the methane over the period of melting. There are a number of such sources. For instance:
Methane seeps from coal measures, shales and oil deposits*. Not only do such formations contain considerable methane but as the ice sheet pushed down on the continent, this pressure would have put strain on underlying rocks and possibly opened up cracks, allowing methane to escape in sort of a natural fracking. Such methane escape happens all the time when there is no ice cover but the carbon is incorporated into the biosphere as it enters the air. It is then available for sequestration in various sinks. However, with an ice cap, all this carbon would accumulated over the duration of the ice cap to be released suddenly when the ice melts. A hundred thousand years of geological methane seep could amount to a considerable amount of carbon ready to be be released rather suddenly.
* Here is a quote from an article on the work of Katey Walter Anthony, a scientist working in the University of Alaska on the methane which is observed coming out of the land.
"During ground surveys, they examined the chemical and isotope composition of the bubbling methane to determine where it was coming from. In many of the smaller bubbling seeps methane was newer, formed when plants and other organic material decayed in the lakes. However, they found that the largest seeps were outgassing fossil methane from ancient sources, such as natural gas and coal beds. Much of the seeping geologic methane had been trapped underground for tens of thousands of years, meaning that permafrost was thawing to such an extent that it was finally releasing those long-stored gases."
A second source of methane is the decomposition of organic material. Unlike a valley glacier which is constantly moving down a valley and scraping the rock bare underneath it, a continental glacier just sits on the land until it is so thick that it starts to be squeezed outward. A lake, a swamp or a thick deposit from a tundra can be capped and if the land is reasonably flat, there will be little if any horizontal movement of the bottom layer of ice relative to the underlying land. With no contact with the atmosphere, oxygen in these organic rich environments will be quickly used up and anaerobic methanogenesis will start. Over the hundred or so Milena that the ice cap is extant, all this methane should be accumulated as clatrates at the bottom of the ice sheet.
Incidentally, this may explain a possible carbon source to help trigger the melt of the glaciers. Once the ice is thick enough, it flows like taffy. At the outer edges of the ice sheet, the ice is moving horizontally with respect to the ground. Once the ice is thick enough, there could be considerable outflow of methane from the bottom of the ice sheet. A Milankovitch nudge might be just enough to tip the balance.
It is interesting to note that Carbon dioxide also produces a clathrate under similar conditions, so any source of Carbon dioxide being released from the earth under the ice would also likely form a clathrate which would be released as a continental glacier melted. The formula for CO2 clatrate is CO2.6H2O. Sources that could release Carbon dioxide are basically only volcanic action. Any disintegrating organic material would quickly shift to producing methane as soon as the residual oxygen had been used up.
It might be possible to detect methane or carbon dioxide when ice coring in Greenland or Antarctica reaches bedrock. A hole to the bottom of the ice 'transmits' one atmosphere pressure to the bottom of the hole and both methane and Carbon dioxide clathrates break down and give up their gas at atmospheric pressure. It also might be possible to detect methane or carbon dioxide at the edges of today's ice sheets.
A further source would be permafrost. At present is it believed that considerable methane clathrate is stored in permafrost. As odd as it seems, a cover of ice insulates this permafrost from sub zero air and geological heat then begins to melt the permafrost from below. It would not be expected to find permafrost under an old ice cap. The warming of the permafrost would liberate its store of methane which would then accumulate at the bottom of the ice sheet, ready to be released when the ice sheet melted and thus contribute to a run away feed back loop.
There is a body of opinion amongst scientists that if the tropical oceans of the world warm up by only a few degrees, corals will eject their zooxanthellae, stop growing and die. Whether or not this will occur remains to be seen (probably fairly soon). If so, it could be a further reason for the observed rise in atmospheric CO2 at the start of interglacials. Just like the suppression of ocean-overturn, the cessation of coral growth is the shut down of a sink rather than the start or increase of a source. The skeletons of coral are made of Calcium Carbonate; (Calcium Oxide and Carbon dioxide). Calcium carbonate is 60.6% Carbon dioxide by weight. Corals and any other marine organism that makes a skeleton of Calcium carbonate, sequester carbon dioxide from the environment. So why would tropical waters be warming up as an interglacial started.
This is related to the shut down of the ocean circulation. The Gulf stream, which warms Britain and Northern Europe also cools tropical areas. Without this constant flow of heat northward, tropical waters would be expected to become warmer. Oddly enough, this could occur just as the seas are getting deeper and the possibility opens up for corals to undergo a huge growth spurt due to the surface of the ocean no longer constraining their growth.
In the coral record from the end of this latest glacial period, one might see a check in coral growth and possibly a change to species that grow better in slightly deeper water followed by a strong upsurge in growth as soon as the ice has all melted and the Gulf Stream re-established itself.
Note that once the ice has all melted and the growth of corals starts again, the potential uptake of CO2 is immense. The sea will have risen a hundred meters or so* and the corals will then grow back up to the surface of the sea. At present corals extend right up to the low tide level. If the 125,000 years of the last glacial resulted in the corals being eroded to sea level, all this top layer of 120m of coral has grown since the last ice age finished 11,500 years ago.
* 120 meters since the end of the last glacial and still rising
In locations where coral material is more than 120m deep, there should be an age discontinuity at about 120m. This would be visible using the uranium dating method.
In summary, we have Four possible sources of the observed spike in Carbon dioxide at the beginning of an interglacial as continental ice sheets start to melt. These are:
a)the upsurge of ice-suppressed volcanism,
b)shut down of oceanic circulation and hence photosynthesis,
c) the release of accumulated clathrates under the ice and
d)the shut down of coral growth.
Each of these would create a feed back global warming which would further encourage the melting of the ice. Once all the ice had melted, carbon dioxide sinks would once more remove this gas from the atmosphere and we would gradually head toward another glacial.
We are now seeing a further act in this saga. It seems likely that within a decade or two, due to man's frantic rush to put sequestered carbon back into the atmosphere, there will be a virtually ice free Arctic ocean. The Arctic ocean then becomes a huge solar panel absorbing heat from the sun. Just on the edge of this ocean on Greenland is the last remnant of the northern hemisphere continental glaciers. If we have sudden melting of this mass of ice, we may see, in miniature, the repeat of the end of an ice age. It should be interesting.