Has the puzzle of rapid climate change in the last ice age been solved?
Bremerhaven, Germany, August 19th, 2014. Over the past one hundred thousand years cold temperatures largely prevailed over the planet in what is known as the last ice age. However, the cold period was repeatedly interrupted by much warmer climate conditions. Scientists have long attempted to find out why these drastic temperature jumps of up to ten degrees took place in the far northern latitudes within just a few decades. Now, for the first time, a group of researchers at the Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research (AWI), have been able to reconstruct these climate changes during the last ice age using a series of model simulations.
The surprising finding is that minor variations in the ice sheet size can be sufficient to trigger abrupt climate changes. The new study was published online in the scientific journal Nature last week and will be appearing in the 21 August print issue.
During the last ice age a large part of North America was covered with a massive ice sheet up to 3km thick. The water stored in this ice sheet is part of the reason why the sea level was then about 120 meters lower than today. Young Chinese scientist Xu Zhang, lead author of the study who undertook his PhD at the Alfred Wegener Institute, explains. “The rapid climate changes known in the scientific world as Dansgaard-Oeschger events were limited to a period of time from 110,000 to 23,000 years before present. The abrupt climate changes did not take place at the extreme low sea levels, corresponding to the time of maximum glaciation 20,000 years ago, nor at high sea levels such as those prevailing today – they occurred during periods of intermediate ice volume and intermediate sea levels.” The results presented by the AWI researchers can explain the history of climate changes during glacial periods, comparing simulated model data with that retrieved from ice cores and marine sediments.
How rapid temperature changes might have occurred during times when the Northern Hemisphere ice sheets were at intermediate sizes
During the cold stadial periods of the last ice age, massive ice sheets covered northern parts of North America and Europe. Strong westerly winds drove the Arctic sea ice southward, even as far as the French coast. Since the extended ice cover over the North Atlantic prevented the exchange of heat between the atmosphere and the ocean, the strong driving forces for the ocean currents that prevail today were lacking. Ocean circulation, which is a powerful “conveyor belt” in the world’s oceans, was thus much weaker than at present, and consequently transported less heat to northern regions.
During the extended cold phases the ice sheets continued to thicken. When higher ice sheets prevailed over North America, typical in periods of intermediate sea levels, the prevailing westerly winds split into two branches. The major wind field ran to the north of the so-called Laurentide Ice Sheet and ensured that the sea ice boundary off the European coast shifted to the north. Ice-free seas permit heat exchange to take place between the atmosphere and the ocean. At the same time, the southern branch of the northwesterly winds drove warmer water into the ice-free areas of the northeast Atlantic and thus amplified the transportation of heat to the north. The modified conditions stimulated enhanced circulation in the ocean. Consequently, a thicker Laurentide Ice Sheet over North America resulted in increased ocean circulation and therefore greater transportation of heat to the north. The climate in the Northern Hemisphere became dramatically warmer within a few decades until, due to the retreat of the glaciers over North America and the renewed change in wind conditions, it began to cool off again.
“Using the simulations performed with our climate model, we were able to demonstrate that the climate system can respond to small changes with abrupt climate swings,” explains Professor Gerrit Lohmann, leader of the Paleoclimate Dynamics group at the Alfred Wegener Institute, Germany. In doing so he illustrates the new study’s significance with regards to contemporary climate change. “At medium sea levels, powerful forces, such as the dramatic acceleration of polar ice cap melting, are not necessary to result in abrupt climate shifts and associated drastic temperature changes.”
At present, the extent of Arctic sea ice is far less than during the last glacial period. The Laurentide Ice Sheet, the major driving force for ocean circulation during the glacials, has also disappeared. Climate changes following the pattern of the last ice age are therefore not to be anticipated under today’s conditions.
“There are apparently some situations in which the climate system is more resistant to change while in others the system tends toward strong fluctuations,” summarises Gerrit Lohmann. “In terms of the Earth’s history, we are currently in one of the climate system’s more stable phases. The preconditions which gave rise to rapid temperature changes during the last ice age do not exist today. But this does not mean that sudden climate changes can be excluded in the future.”
The paper:
Xu Zhang, Gerrit Lohmann, Gregor Knorr, Conor Purcell:
Abrupt glacial climate shifts controlled by ice sheet changes.
Nature, DOI: 10.1038/nature13592
During glacial periods of the Late Pleistocene, an abundance of proxy data demonstrates the existence of large and repeated millennial-scale warming episodes, known as Dansgaard–Oeschger (DO) events1. This ubiquitous feature of rapid glacial climate change can be extended back as far as 800,000 years before present (bp) in the ice core record2, and has drawn broad attention within the science and policy-making communities alike3. Many studies have been dedicated to investigating the underlying causes of these changes, but no coherent mechanism has yet been identified3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15. Here we show, by using a comprehensive fully coupled model16, that gradual changes in the height of the Northern Hemisphere ice sheets (NHISs) can alter the coupled atmosphere–ocean system and cause rapid glacial climate shifts closely resembling DO events. The simulated global climate responses—including abrupt warming in the North Atlantic, a northward shift of the tropical rainbelts, and Southern Hemisphere cooling related to the bipolar seesaw—are generally consistent with empirical evidence1, 3, 17. As a result of the coexistence of two glacial ocean circulation states at intermediate heights of the ice sheets, minor changes in the height of the NHISs and the amount of atmospheric CO2 can trigger the rapid climate transitions via a local positive atmosphere–ocean–sea-ice feedback in the North Atlantic. Our results, although based on a single model, thus provide a coherent concept for understanding the recorded millennial-scale variability and abrupt climate changes in the coupled atmosphere–ocean system, as well as their linkages to the volume of the intermediate ice sheets during glacials.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Findings? In a computer model? These guys really do believe the “puter” is the new laboratory, which it ain’t!
The climate in the Northern Hemisphere became dramatically warmer within a few decades until, due to the retreat of the glaciers over North America and the renewed change in wind conditions, it began to cool off again.
Uh, what caused the “change in wind conditions”?
Ice doesn’t melt just because it feels like melting; there must be an underlying cause for the ice to melt. Wouldn’t that cause be the reason for the climate to change?
As a result of the coexistence of two glacial ocean circulation states at intermediate heights of the ice sheets, minor changes in the height of the NHISs and the amount of atmospheric CO2 can trigger the rapid climate transitions via a local positive atmosphere–ocean–sea-ice feedback in the North Atlantic.
They just had to stick that in for the grant thingy.
At this point, my opinion is that humans understand about 1% of what drives climate changes over time periods longer than a year. Does anyone believe that our knowledge is more solid than that?
Caleb, your overturning (or simply ‘turning’) is why the city folks who get their water from lakes suffer with water that tastes and smells like dirt several times a year. As our area switched from wells to lakes as our water supply, the difference was spectacular. And not in a good way.
“Small variations in the climate system can result in dramatic changes in temperature”
1) Sound like the definition of a non-linear chaotic system. …. and so why would we expect a log-linear system with CO2 changes ?
2) Also sounds like the Lorenz’s Butterfly effect – theory developed in 1961 – 53 years ago. I am not sure why the above quote would be surprising to anyone in the field, unless they are completely unaware of the literature
I wish that researchers (and reporters) would stop using the term “ice age” when they are clearly referring to a “glacial epoch.” An “ice age” consists of several glacial epochs interspersed with interglacial periods. We are currently in an ice age. The previous ice age occurred 150 million years ago. These researchers should know better.
I am only listing one of the 4 which applies to this particular article which I have been addressing al along which is when the climate is close to a glacial/interglacial situation abrupt climatic changes are more likely to occur with GIVEN forcing.
This will be my only commentary limiting myself to one per day per topic.
MY FOUR FACTORS
1. The initial state of the global climate.
a. how close or far away is the global climate to glacial conditions if in inter- glacial, or how close is the earth to inter- glacial conditions if in a glacial condition.
b. climate was closer to the threshold level between glacial and inter- glacial 20,000 -10,000 years ago. This is why I think the climate was more unstable then. Example solar variability and all items would be able to pull the climate EASIER from one regime to another when the state of the climate was closer to the inter glacial/glacial dividing line, or threshold.
They had me interested right up until the last line of the first paragraph:
using a series of model simulations
Jeff L. says:
I am not sure why the above quote would be surprising to anyone in the field, unless they are completely unaware of the literature
Actually, unless they are completely uneducated on the subject in general!
One should not view the AMO as a closed system or primary driver of weather pattern changes. An ENSO signal arrives about 6 months later and shows up in the AMO data, rather consistently and noisily. Here is a paper explaining that ENSO signature in the Atlantic Multidecadal Oscillation Index (AMOI) data and suggests a 6 month lead time, meaning that the Multivariat ENSO Index (MEI) demonstrates a 6 month lead to that signal showing up in the Atlantic. The teleconnected forcing is likely an atmospheric bridge:
http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&uact=8&ved=0CCgQFjAB&url=http%3A%2F%2Fwww.esrl.noaa.gov%2Fpsd%2Fpeople%2Fgilbert.p.compo%2FCompoSardeshmukh2008b.pdf&ei=JAj2U9OyOuS8igL5hoG4Cw&usg=AFQjCNFtMFvtLLNqno1X7llx3B9kr8WRqQ&sig2=EYxYPOFNdKZ-lyWpbGJlEQ&bvm=bv.73231344,d.cGE
I love the following 2010 paper about the impossible. As in removing the ENSO signal from the larger SST data on a global basis. The authors throw up their hands in the end and give in to the idea it can’t be done. It can’t be modeled and it can’t be removed. The AMO may be a darling child to study, but it is likely a child and not the adult. The adult in this mess is most likely the granddaddy of SST variation and trends: ENSO.
http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=10&cad=rja&uact=8&ved=0CG8QFjAJ&url=http%3A%2F%2Fwww.esrl.noaa.gov%2Fpsd%2Fpeople%2Fgilbert.p.compo%2FCompoSardeshmukh2008b.pdf&ei=2AX2U4WjNeTUigKPkoHADw&usg=AFQjCNFtMFvtLLNqno1X7llx3B9kr8WRqQ&sig2=0DD3w8fGBFz1XzkSS5Goiw&bvm=bv.73231344,d.cGE
But you will love the unexpected ending. There are plenty of papers that propose that the AMO then triggers changes in the North Pacific.
http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&uact=8&ved=0CCYQFjAB&url=http%3A%2F%2Fgfdl.noaa.gov%2Fbibliography%2Frelated_files%2Froz0704.pdf&ei=2Qr2U9jbMsmaigKigYFY&usg=AFQjCNFgh4S4QtV0X3ZYzwU8ylYpIby95A&sig2=5qGPQkmGZuCMwq0guCK0xg
And the two together, the PDO and the AMO, can wreak havoc with a decades worth of Sunday afternoon outings.
http://oceanworld.tamu.edu/resources/oceanography-book/oceananddrought.html
Scientists have long attempted to find out why drastic temperature jumps of up to ten degrees take place in the far northern latitudes within just a few hours.
Fixed it
A big problem with computer models is that the results depend not only on what assumptions you put into the model, but also what you don’t put in.
Some major problems with this particular model include:
1. “The rapid climate changes known in the scientific world as Dansgaard-Oeschger events were limited to a period of time from 110,000 to 23,000 years before present. The abrupt climate changes did not take place at the extreme low sea levels, corresponding to the time of maximum glaciation 20,000 years ago, nor at high sea levels such as those prevailing today – they occurred during periods of intermediate ice volume and intermediate sea levels.”
This statement is dead wrong–both the oxygen isotope ratios and temperature curves from the GISP2 ice core show that the most sudden warming that brought the last Ice Age to a close occurred abruptly 15,000 years ago when ice sheets were at their maximum extent and sea level was at its lowest. The ice sheets had been at their late glacial maximums for several thousand years when, out of the blue, temperatures suddenly soared 13 C (23 F) in something like 100 years, causing wholesale melting of the ice sheets. Then, just as abruptly, temperatures turned around and cooled 10 C (18 F) by 14,000 years ago. Temperatures then fluctuated up and down( but not so intensely) at intermediate levels for about 1,000 years. 12,700 years ago, temperatures took another nosedive into the Younger Dryas cold period and remained at full glacial conditions for 1,000 years. During the Younger Dryas, temperatures repeatedly changed abruptly from cool to warm (the Dansgard-Oeschger events). 11,500 years ago, another great warming spike caused temperatures to soar 12 C (21 F) in about 100 years (at one point, around 20 degrees in 40 years), then continued warming at slower rates for a total warming of 17 C (30 F) from 11,700 to 10,000 years ago. (Keep in mind that these temperatures are for Greenland, not global, but they correlate very well with temperature conditions in the rest of the world). These abrupt, multiple, intense changes (back and forth) at full glacial conditions (not ‘intermediate’) hardly sound like changes caused by gradual changes in ocean/atmospheric conditions.
2. All of the five most significant colder temperature changes that occurred during the past 500 years (the Little Ice Age) coincided with low sunspot intervals, lower total solar irradiance, lower solar magnetic flux, and increases in the production rates of beryllium-10 and carbon-14 (isotopes created in the upper atmosphere by increase in cosmic ray flux rates). These are not mere coincidences–such good correlation is not random chance, but must be due to cause-and-effect circumstances. None of this, of course, made it into the computer model simulations.
There are other serious problems with the methodology in this paper, but the bottom line is that (1) their initial premise of changes occurring only during ‘intermediate’ glacial conditions is dead wrong, and (2) sunspot intervals, lower total solar irradiance, lower solar magnetic flux, and increases in the production rates of beryllium-10 and carbon-14 argue strongly for other factors not included in their models. Thus, their conclusions cannot be considered valid.
Looking at their figures, ( http://www.nature.com/nature/journal/v512/n7514/full/nature13592.html ) they do not seem to get the sea level correct. They forgot that all water depths 120 metres down were land surfaces. So no Baltic sea, no water cooling the N. Pacific thru the Bering Straight, no Hudson Bay and all the islands of N Canada had no water currents. Also the seem to not realize how much Arctic Ice melt is due to warmer fresh water run off from the land. With glaciers all around, no run off.
Read a thought provoking article where the author (I forgot who) questioned whether if the deep ocean heating would go up because of the undersea volcanoes, do to the ice cap preventing evaporative heat loss, like a pot boiling with or without a cover.
With a chaotic, multivariate complex system model of the climate, even if it is perfect, you might be able to demonstrate rapid climate change but you can never predict it (incomplete knowledge of initial conditions and insufficient computing power). Therein lies the rub and so the IPCC’s scenarios instead of predictions. I think the best bet is to observe patterns and project ahead (e.g. the Holocene is cooling down and we’re heading into a glaciation). All the fancy model stuff does is give your guesses a patina of scientism with which to fool the funding agencies.
The temperature record is really a record of ice sheet extent: land ice plus sea ice, land ice being about a thousand times thicker than sea ice and requiring a thousand times longer to melt. Ice melts on clear summer days, and gets an extra 80W/m^2 when orbital conditions are right. Albedo feedback is on the order of 30W/m^2 at peak insolation. GHG feedback is 1 or 2 watts, 24/7, which means little on a winter night over ice. Ice sheets create their own climate, which might be expected to replicate that of the Antarctic to some extent; at any rate, hard to model.
Clearly Milankovitch cycles trigger ice ages, where land ice responds at periods in the same ballpark, lagging by thousands of years. Over the long term sea ice grows and melts in tandem with the big ice sheets, but sea ice can respond on short order to short term influences and create sudden changes in average global temperature as recorded in the ice cores. Secondary positive feedback loops of sea ice would be a thousandth as long as those of land ice. –AGF
Don Easterbrook says:
August 21, 2014 at 8:41 am
You are correct. Fluctuations occur on about the same period during glacial maximum, normal glacial and interglacial conditions. During glacials, they’re called D/O events and Bond cycles during interglacials. The size of the fluctuations is greater during glacials because the moves start from such low temperatures.
During glacials there are also Heinrich events at fairly regular intervals.
“Secondary positive feedback loops of sea ice would be a thousandth as long as those of land ice.” –I sure got that backwards! –AGF
Caleb,
I’m not a scientist, but a pond in which there is no mixing current is very different than an ocean in which there is. Ocean heat moves by both conduction and convection. Ponds without anything to move the water layers around, can only move heat by conduction. Cold water does not “rise” to the top of a pond. The contact with the air that has become colder than the water at the surface causes the water at the surface to lose heat to the air as the two attempt to reach equilibrium. The thermal insulating properties of the land at the bottom and sides of the pond allow the water beneath the surface to hold onto its warmer temperatures longer than the water at the surface, which is in contact with the air. Air both warms AND cools faster than both water and land. Thus the water in contact with it will cool faster than the warmer layers beneath the surface can adjust for the temperature differences between them and reach thermal equilibrium with each other.
“Many studies have been dedicated to investigating the underlying causes of these changes, but no coherent mechanism has yet been identified”
We really don’t know.
We have plenty of experts in various related fields that think they know. They present/sell their theory with a high level of confidence and with compelling facts. Proxy data is great but it can/will send you down the wrong path at times.
If the measurements line up consistently a certain way, it’s a clue but there is still much more missing pieces to the puzzle than their are discovered pieces.
Over the last 100 years, our ability to actually measure the atmosphere, oceans, sun and related elements gives us the ability to see everything as it is actually happening………..and we are still missing some key pieces needed to solve the puzzle.
Regarding pieces of the puzzle we find that we think are from 100,000 years ago or millions of years ago. They can tell us with high confidence about something that happened in the past. We can guesstimate the time frame but could be off.
I think that sometimes, we might be trying to fit those pieces into the wrong puzzle. An even more powerful force, with no evidence left behind could be entirely responsible for causing what we think our only evidence must have caused.
We can’t be sure about any changes that involved our relationship with the sun or other things, that even right now, we are trying to figure out.
Our biggest limiting factor is that the real key to the entire puzzle requires capturing/measuring changes as they occur and they occur on a geological time scale, with humans only doing this for a tiny fraction of time on that scale.
The sun may be obliging to some extent at this fraction in time, with it’s recent cycle changing enough for us to compare it to previous cycles.
Don Easterbrook,
I sent you an email last week to ask a question about whether or not it was possible to tie the “new” pingos to the explosion/impact of the Chelyabinsk meteor last year. Researchers discovered that the ultrasonic sound and percussion waves were powerful enough to be detected by seismographic instruments around the world. They were detected circling the planet TWICE before dissipating. If there was that much subterranean disturbance/energy in play, is it possible that it could have loosened ice cores in the pingos enough to cause them to burst ? As remote as these locations are, for them to not be noticed for a year isn’t unusual. Any way to check satellite images from just prior and after impact for comparison?
So is it minor variations in ice sheet size that trigger abrupt climate changes, or is it abrupt climate changes that trigger variations in ice sheet size? How do they tell the cause from the effect? Do ice sheets vary in size on their own without any changes occurring in the climate first?
Louis says:
August 21, 2014 at 11:31 am
Comparing June insolation at 65°N with δ18O here: http://en.wikipedia.org/wiki/Milankovitch_cycles#mediaviewer/File:Vostok_420ky_4curves_insolation.jpg
leaves no room for doubt as to the cause: the earth’s orbit; not CO2, not CH4. It is impossible for functions of correlating periodicity to have anything but a common cause. If you lived in the tropics you could be oblivious to ice ages, but where lasting ice appears or disappears there is no separating climate from ice extent, or local temperature, or global temperature. –AGF
PS, I guess I didn’t get that last line backwards.
Clearly the Gulf Stream turns south. This winter sea ice in the Arctic will quickly was incremental.
http://earth.nullschool.net/#current/ocean/surface/currents/overlay=sea_surface_temp/orthographic=-46.10,50.00,837
“There are apparently some situations in which the climate system is more resistant to change while in others the system tends toward strong fluctuations,” summarises Gerrit Lohmann. “In terms of the Earth’s history, we are currently in one of the climate system’s more stable phases. The preconditions which gave rise to rapid temperature changes during the last ice age do not exist today. But this does not mean that sudden climate changes can be excluded in the future.”
Uh-Huh….