You wouldn’t think that the direct route from Nairobi to Seattle goes over Greenland.  But just try stretching a string over a globe’s surface if you don’t believe me.  The shortest string goes right over the Sudanese and Libyan deserts, then over Copenhagen, northern Iceland, and Greenland.

     I particularly like the view from the great circle route connecting Copenhagen to Seattle, so I have managed to get a “geologist’s seat” (window seat, in front of the wing, on the shady side of the plane, take your binoculars).  One flight in three, the clouds part and make your efforts worthwhile.

     From Copenhagen, the plane goes over Norway and then the Greenland-Iceland-Norwegian Sea (sometimes called the “Nordic Seas” but I’ll just call it all the Greenland Sea).  Next it passes over a mountainous minor continent of rock and ice paradoxically called Greenland (a name devised by Eirik the Red to promote immigration, an early instance of deceptive advertising).  Then we go over the great stores of methane frozen into the Canadian tundra (and apt to be released into the atmosphere by greenhouse warming, making things even warmer still).  Those sites are all key players in climate flip-flop scenarios.  They are intimately connected to our African past and our worldwide future.

 

We’ve talked a lot about abrupt climate jumps without mentioning very much about how they happen.  The big scientific questions often have three major facets:  what, how, and why.  And before I can get much further into the how and why of things, I’d better flesh out the rest of the What’s, the slow background changes which made the abrupt coolings such a surprise to the scientific world in the 1990s.

     It seems strange to realize that even in the middle of the 19th century, hardly anybody knew about the ice ages.  Even scientists still talked in terms of a biblical deluge.  But as early as 1787 in Scotland, scientists realized that giant boulders were carried long distances and deposited in the midst of a landscape in which their kind of rock was utterly foreign.  Even peasants in Switzerland subscribed to such an explanation.  Charles Darwin, in his Voyage of the Beagle, wrote in 1839 of boulders carried along in icebergs.  Louis Agassiz  spent a lot of time looking at alpine landscapes.  Those Swiss valleys looked as if they had been scoured, over and over.  From his position as president of a Swiss scientific society in 1837, Agassiz elaborated the earlier notions and proposed that massive ice sheets had pushed their way around Europe, sometime in the not-so-distant past.  Darwin provided further geological details on the subject in 1842.

     The landscape evolved.  Not only did the Alps grow long glaciers but giant ice sheets sat atop much of Northern Europe, flattening the Baltic landscape.  Some piles of ice didn’t even have mountains centered beneath them; they were like sand dunes, obstacles that grew as the laden winds ran into them.

     We eventually learned that the warm periods in the ice ages last only about 10,000 years and that it’s a long time until our next one is “due” – about 100,000 years.  We learned that 32 percent of the land was covered by ice sheets in the last ice age, but that a warm period like today’s reduces the ice to only 10 percent (mostly Greenland and Antarctica, but you can also see some serious ice sheets in Norway, Iceland, Alaska, Tibet, New Zealand, and the Andes).

     How many ice ages have there been?  Dozens.  Up until about 800,000 years ago, the major meltbacks occurred about every 41,000 years.  More recently, the 100,000 year rhythm has been the more obvious one.

The second great insight concerned how sunlight might vary to help melt off the massive amounts of ice.  These slow changes in ice mass are mostly caused by even slower changes in the summer sunshine at latitudes like Scandinavia.

     Elaborating on the 1842 suggestion of the French mathematician Joseph A. Adhémar, Milutin Milankovitch proposed during World War I that the sunlight reaching the higher latitudes controlled the ice ages (he did his laborious calculations as a prisoner of war), and that slow changes in the earth’s orbit were important.

     First of all, the tilt of the earth’s axis varies over a 41,000 year cycle.  Our axial tilt peaked 9,500 years ago at 24.6° and is now at 23.4° and heading toward a turnaround at about 22.1°.  Furthermore, our closest approach to the sun in our elliptical orbit is currently in the first week of January (we’re about 3 percent nearer, and get about 10 percent more heat, than in July) – but in another 12,000 years or so, the closest approach date will have drifted around to summer again, the precession helped by the gravitational pull of the other planets causing our less-than-spherical planet to slowly wobble.  Our orbit is also more circular at some times (repeating at about 100,000 and 400,000 years), making the month-of-closest-approach factor periodically less important.

     Tilt and precession account for a lot of the slow back and forth of the glaciers in between the major meltoffs, which occur every 100,000 years or so.  The best setup for melting ice is when there are colder winters but hotter summers in the Northern Hemisphere (that is, after all, where most of the meltable ice is, thanks to the general lack of land between 40°S (southernmost Australia) and Antarctica.  (Yes, I know, there are the southern Andes and some nice glaciers in southern New Zealand, but that’s not much when compared to the land mass of Eurasia and North America north of Madrid and Kansas City.)

     “Exaggerated seasonality” occurs when the closest-approach bonus is in summer and there’s also another bonus from near-maximum tilt.  In such cases,  there is as much summer sunlight every day at 65° (say, in northern Iceland or Fairbanks, Alaska) as there is presently at 49° (say, in Paris or Victoria).  Despite the accompanying colder winters, getting melting going during those long hot summers is how we got rid of the ice sheets at high northern latitudes.  At the opposite extreme, 65° gets about what Thule (78°N) gets today.  These rhythms, and the way they reinforce one another occasionally, also explain a lot of the back-and-forth between one major meltoff and the next, the so-called interstadials.

 

But it was long suspected that “orbital factors” weren’t the whole story, as the southern hemisphere ice sheets melted back at the same time as the northern ones.  Why should ice sheets in the Andes and New Zealand melt when the closest-approach bonus was in their wintertime?  They should have been out of phase with the northern one, but they were in phase, synchronized by something.

     The other thing that made scientists suspect that it wasn’t so simple was a wildflower, white with a yellow center, of the rose family called Dryas octopetala.  It is cold-adapted and is found on the tundra – not among shrubs and pine trees.  Well, a century ago, Dryas pollen turned up in cores of a lake bed in Denmark above a layer of trees, at a depth now dated back to about 12,000 years.  A return to cold made no sense then, according the Milankovitch orbital view.  That was when the two astronomical bonuses were combining to produce particularly hot summers in Denmark.  Half of the ice sheets covering Europe and Canada had already melted, and all of the ones in Scotland.  There should have been pine seeds in those cores, not the pollen of a tundra flower.

     This return to ice age temperatures, called the Younger Dryas, lasted more than a millennium, reestablishing glaciers in Scotland once again.  Then, things suddenly warmed back up.  Nothing in the orbits could explain such things.

 

If I had visited Copenhagen researchers in the late 1970s, no one would have agreed on when the ice ages began.  The textbooks, noting the age of old moraines plowed up by the ice sheet frontiers, would have said the ice ages went back about 800,000 years.  No one yet knew about winter sea-ice reaching the British Isles 2.51 million years ago, nor about all those antelope speciations in Africa about the same time.

     Most people thought, back then, that the ice buildup and the ice melting were largely due to those slow changes in various aspects of the earth’s orbit around the sun.  Since they don’t suddenly jump, most people assumed climate would change gradually.  The cores coming out of the ocean floor indeed showed slow changes, and so did the ice core from Antarctica – but few understood then about how special Antarctica is, how insulated from the rest of the world its weather can be (that ring of westerlies and its vertical curtain of rising air at 60°S) and how its low rates of snowfall limit time resolution.  (Various parts of Antarctica are now known to give different results.)

     But the pros knew about such anomalies as the Younger Dryas and they were anxious to get better data.  Greenland had more snow each year, thanks to weather systems sweeping up from the Gulf Stream, and the Greenland cores were starting to show some puzzling results, as the Danish researcher Willi Dansgaard and colleagues reported in 1982.  I first heard of the abruptness, per se, in 1984 when the Swiss geophysicist Hans Oeschger gave a talk in Seattle.  The time calibration on one of his slides prompted me to ask him afterward, about just how quickly temperature had changed.

     Oh, he said, the big drop took just a few years.  The enormity of such a whiplash caused me to assume that we were having some language difficulties and so I persisted, asking, “Just a few decades?”  No, no, he replied, merely a few years.

     The American geochemist Wallace Broecker also heard Oeschger give a talk in 1984, and the quick flips gave Broecker his idea for different modes of circulation via failures of the salt conveyor (more in a minute).  It was Broecker who coined the term “Dansgaard-Oeschger events” to describe the abrupt coolings and re-warmings.  (The D-O’s are what I’ve been calling the flip-flops or abrupt jumps).

     Another important discovery was from the ocean-floor cores of the North Atlantic.  Now and then, a rain of rock had fallen down to the abyss, dropped off the bottom of passing icebergs.  The rocks mostly came from Hudson Bay (though some came from Iceland).  These “Heinrich events” are not completely understood.  They do tend to occur in the coldest parts of the glacial cycles, when ice sheets extend out on to continental shelves.  Perhaps they represent glacial advances that overrun terminal moraines and then freeze their rock into the bottom of the ice sheet.  Later, when tides manage to float a terminal glacier and break it off at a weak “hinge,” an iceberg is created which sets sail for the Bay of Biscay, held upright by its rocky ballast on the bottom.  As it melts en route, so a trail of debris is deposited on the ocean floor – and fresh water is leaked all over the surface of the mid-Atlantic Ocean.  That the events are episodic, lasting for several centuries, is attributed to the Hudson Bay ice mountain collapsing, perhaps from melting on the bottom in the manner of subglacial lakes in Antarctica.

     The D-O flips are not the same thing as the Heinrich events, though the issue was confused for awhile.  For example, a D-O cooling occurred in the midst of the last warm period at about 122,000 years ago, when there were no ice sheets in Canada to create icebergs.  But in icy times, Heinrich events do tend to be shortly followed by an abrupt D-O warming.

 

Summer is drilling season and Greenland is where most of the Copenhagen researchers currently are.  They try to extract long cylinders of ice, ones that will show annual layers of snow and ice.  Greenland has ice that goes back to about 250,000 years ago, when it was warm for long enough that it melted most of the accumulation from prior ice ages.

     Still, that quarter-million years seen in the ice cores contains the last two ice ages.  And it includes the last warm period between 130,000 and 117,000 years ago, when there was a climate somewhat warmer than today’s for much of the 13,000 year period.  (For Europe, this warm period is called “The Eemian” – more generally, “marine isotope stage 5e.”)  Sea level was 3-5 meters higher than at present (just imagine shrinking southern Florida).  For comparison, complete melting of the West Antarctic ice sheet (or, for that matter, the Greenland ice sheet) would today raise sea levels by 6–7 meters, which is several stories high.  Melt all the ice sheets and it’s more like 60 meters, covering up 18-story Florida condos, with the coastline somewhere up in Georgia.

     The last ice age per se had several dozen D-O cycles, a flip one way or the other occurring on average about every 750 years.  Our current warm-up, which started about 15,000 years ago, began abruptly in the Northern Hemisphere, with the temperature rising sharply while most of the ice was still present.  Several thousand years later, temperature declined abruptly into the Younger Dryas.  After the sudden rewarming at 11,550 years ago that ended the Younger Dryas, things gradually warmed into the period of modern sea level and agriculture.

     There was a slow-and-gradual view of things in the good old days, back before we got the time resolution to see the abrupt stuff that our ancestors suffered through.  These slow “orbital” causes of ice fluctuation have little to do with the rapid back-and-forth of a whiplash.  Something else seems to be the more immediate cause of the chattering seen atop the slower orbital trends of ice.

     Because some of the D-O cold-and-dry flips last for many centuries, there is a semi-slow way of viewing them, and indeed a group of European researchers are looking at the cluster of flips between 60,000-25,000 years ago for clues about the demise of the Neandertals, calculating the expected regional climate changes in Europe and correlating them with the dating of archaeological sites to see how the Neandertal populations were moving around in response (see their database web pages at http://www.esc.cam.ac.uk/oistage3/ ).

     But my bust-then-boom aspect is keyed to the initial part of the abrupt changes, just the first century or two after a cooling-and-drying (when the drought-to-fire-to-grass-to-herds opportunity occurs) and the century after the abrupt rewarming that ends it (when grasslands extend into formerly arid areas, and again dramatically expand herd sizes).  That first century after a flip should be when unusual opportunities present themselves for experienced hunters to expand their territory and populations.  Archaeological time resolution cannot, at present, see such decade-to-century transitions, but that’s likely where the action is.