Rough winds do shake the darling buds of May
In the previous posts we looked at the Eifel volcanics, the Laacher See (the largest eruption in western Europe for the past 1000 centuries), and volcanic dating of the Younger Dryas. The eruption occurred a bit over a century before the sudden cooling. But what caused that cooling?
Let’s go back further, to the depth of the ice age long before the Younger Dryas. In the middle of the ice age there were multiple rapid warmings followed by slow cooling. They called Dansgaard-Oeschger cycles (no, this is not the name of a plant); at least 25 of these spikes are recognized in the ice core record. In each spike, local temperatures would rise by as much as 10C. They were most common during the period between 60,000 and 30,000 years ago, but were rare during the coldest phases of the ice age.
The fact that they are seen so well in 18O shows that they started with warming of the ocean. The events also coincided with strongly increased iceberg activity, perhaps caused by the warming water destabilizing the ice pack. The precise cause remains disputed. A popular option is that melt water from the glaciers disrupted the ocean circulation. Another is a change in wind patterns, induced by the height of the glaciers (a 3-km tall ice cap adds a substantial barrier to the air circulation!). Either way, the suddenness of the changes indicates that there was a significant instability in the ice age climate.
The fall of winter
Now is the winter of our discontent
The Younger Dryas resembles a Dansgaard–Oeschger event – but in reverse. Instead of rapid warming, it started with fast cooling. It also lasted longer, and while during a Dansgaard–Oeschger event there is a slow temperature drop during the (typically) few hundred years of (relative) mildness, the Younger Dryas showed a slow re-warming. Both events show a very sudden end, in the case of Dansgaard–Oeschger a temperature collapse back into deep ice age, and in the case of the Younger Dryas an almost instant warming. Really, it was nothing like a Dansgaard–Oeschger event.
The plot shows the comparison between the events. The Younger Dryas as seen in the Greenland ice core record is shown by the bottom curve, as the black dots. The axis is labeled in thousand years. The sharp cooling just after 13,000 b2k is visible, followed by a bit of recovery, and finally a very sharp rewarming.
The lines above it, using plus signs, show three ice age events. Obviously they happened much earlier than the Younger Dryas: they are shown shifted in time. They also were colder: the δ(18O) of these events has been shifted upward. Otherwise they fall on top of the Younger Dryas or below it.
The top two show two long-lasting, one starting around 43,400 and the other one at 38,200 (all in b2k). While most Dansgaard-Oeschger events lasted some 500 years, the longest such events were of similar duration to the Younger Dryas.
The lower of the three lines shows two faster Dansgaard-Oeschger events. The first one begins immediately at the leftmost of the line (which was 34,700 b2k) and ends 700 years later. The second one begins 1200 years later and lasted 400 years. By coincidence, the time between the events looks just like the Younger Dryas: fast cooling followed by rewarming 1200 years later! Perhaps the sudden warming at the end of the Younger Dryas was itself a Dansgaard Oeschger event, where the warming just didn’t end.
The quick repeat of Dansgaard Oescher events was not uncommon during the ice age. I found 10 such ‘pairs’ in the record which were separated by typically 1000-1400 years. This seems to be a common time scale.
Causes of climate change in extremis
Weary of solid firmness, melt itself
Into the sea
The Younger Dryas was an ecological disaster in Europe. Interestingly, the southern hemisphere survived much better. There was no cooling there, and in fact temperatures increased. The Younger Dryas was only in the north. It certainly wasn’t the Sun. It is a bit of a diversion for this post, but it is an irresistable question. What did cause the Younger Dryas and why did it end so sudden?
The favourite explanation for the Younger Dryas relates to a huge lake (‘sea’ is more appropriate) of melt water that had collected at the edge of the North American ice cap during the years of warmer weather. It was contained by a combination of a depressed surface (which was only just beginning to recover from the weight of the glacier) and an ice barrier. Ice barriers are a bad idea in an era of warming. It melted and collapsed, and the lake suddenly emptied into the North Atlantic. Think jokulhaup on a continental scale. Cold fresh water now flooded the ocean. ‘Flooded’ sounds a bit strange (how do you flood something that is already water?) but it is actually the right word. The fresh water was less dense than the salty sea, refused to mix, and remained as a layer of fresh water on top of the ocean of salt – a reverse margarita. It caused chaos.
The North Atlantic is a source of heat for Europe. The salty Gulf Stream brings in the warm water. Here it gives up its warmth to the air and cools. As it cools, the water becomes more dense, and it sinks. This sinking pulls in the next batch from the Gulf Stream. But now there was an insulating fresh water lid on the sea, less dense, which refused to sink. This stopped the Gulf Stream from coming north, and the North Atlantic plunged into deep freeze. 1200 years later, the Gulf stream restarted and the north suddenly rewarmed.
This explains the onset of the Younger Dryas well. But the sudden end has received much less attention. What made the temperature rise so fast, by 10C in a few decades? And why did it last for 1200 years? How could a fresh water lid last so long?
And how about the Dansgaard-Oeschger events which look like an inverse of the Younger Dryas? Were they caused by whatever ended the Younger Dryas, and did they end by with renewed failure of the Gulf stream, in just the opposite sequence of the Younger Dryas?
The answer is most likely in the ocean. But where?
The main suspect is the Labrador current, around Newfoundland. It comes down the west coast of Greenland and floods the ocean with ice-cold water. Fish love it. The Gulf stream has to find its way around it. The idea is that the failure of the Gulf stream after the fresh water flood allowed the Labrador current to expand southward, taking the polar front with it. The Gulf stream found its territory occupied and decided to go to Spain rather than Scandinavia. It took 1200 years before the battle lines shifted again.
There is some evidence that around Newfoundland, there was a slow rewarming of the ocean starting a few hundred years before the end of the Younger Dryas. The Gulf stream was slowly re-invading the occupied territory. At some point it pushed the Labrador current back far enough, re-established dominance, and started flowing north again. The Younger Dryas now suddenly ended.
This puts the cause in the ocean circulation, with a slow strengthening of the Gulf stream. There are other indications for this, which show changes in the tropics around the end of the Younger Dryas.
The main model is that the weakening of the Gulf stream warms the southern Atlantic (south of 20o north), because less tropical warmth is now transported to the far north. This restrengthens the Gulf stream and normal service resumes. But that would predict a relatively fast response, where the cold phase is short-lived. Instead it lasted for 1200 years. And this same time scale shows up in the times between Dangsgaard Oescher events.
There is a possible cause in the deep ocean. The water that sinks in the far north forms a new, icy cold current on the bottom of the sea. It follows the bottom of the Atlantic, skirts Antarctica (where it collects even more cold water), moves east and finally turns back north into the Pacific. In the north Pacific, close to Alaska, it resurfaces, now as a warm current, having picked up heat from the tropics. Another branch does the same in the Indian ocean. The surface current retraces the path and ends up in the Atlantic where it closes the cycle.
This is called the ‘conveyor belt’: it is a major contributor to cycling heat across the globe. The full path takes around a millennium. Close down the cold water input in the North Atlantic, as happened at the start of the Younger Dryas, and the southern and Pacific ocean get less cold water coming to the surface. They become warmer, and send this excess heat back to the Atlantic. After a millennium, like an exceptionally late echo, this heat pulse begins to arrive. It strengthens the Gulf stream, which now begins to push back the invader. It takes a few hundred years before it becomes strong enough to re-occupy the north Atlantic. Sinking resumes, and the Gulf stream again become self-sustaining. Like an ice age Baron von Munchausen, it pulls itself up (or rather, down) by its bootstraps. (In the original story, it was actually by his own pigtail. That never caught on.)
In this model, the cause of the rapid cooling is above water, but the rapid warming comes from the deep, and an ocean away.
GLENDOWER: I can call spirits from the vasty deep.
HOTSPUR: Why, so can I, or so can any man;
But will they come when you do call for them?
The missing eruptions
Let’s go back to the Laacher See. One of the mysteries of this eruption is that it is missing from the Greenland ice record. Other eruptions left their trace of sulphur and sometimes tephra. But no such spike coincides with the Laacher See. How is this possible, seeing how much tephra was distributed over Europe? A VEI 6 in Germany should have shown up there. Was this an accident, with the winds blowing in the most unfavourable location, not even reaching Greenland the long way around? Was the eruption perhaps smaller or less dusty than we thought? (Hunga Tonga, the largest explosion for almost 140 years may leave no trace in the ice because it put its energy mainly into putting water into the atmosphere.) Or did we just look in the wrong place? After all, we had our times wrong for the eruption by more than a century.
The main volcanic marker of the time is the Vedde ash. It dates from midway through the Younger Dryas, is seen across Europe and parts of Russia (deposits in Norway are up to 50cm thick), and came from Katla. The origin was determined from the basaltic and rhyolitic composition. (Yes, that is correct. It produced both, probably within days of each other.) The ash is not present at Katla itself, which was probably thickly glaciated at the time, but it is found in north Iceland, as the so-called Skógar Tephra. There is another rhyolitic ash layer from the Younger Dryas which was found (of all places) in Abernethy forest in Scotland, is a few hundred years younger but has similar composition and is also attributed to Katla.
There are four other ash layers in Greenland which are slightly older than Vedde but date from the Younger Dryas and just before it. One occurred less than a century before the Vedde eruption, was rhyolitic and has the same composition as the rhyolitic component of Vedde. It is also attributed to Katla, as a precursor hickup to the big event.
The next older one is extremely rhyolitic and low potassium. There is no known Icelandic source, but it may have been North American.
An eruption from the early part of the Younger Dryas left a basaltic ash layer. It is called ‘Tv-1’ (no idea why) and is attributed to Grimsvotn.
The last one was shortly before the Younger Dryas and has been seen as the Laacher See deposits. It was at about the right time. However, the chemistry rules it out. It has similarities to Hekla, but this volcano may not have existed at the time.
But most volcanic eruptions in the ice have no tephra in them. Instead we see the sulphate from the emission, which can travel much further than the ash. Below is a plot of sulphate spikes seen in the ice in the time just before the Younger Dryas. This is from Abbott et al, 2021, Quaternary Science Reviews, 274, 107260. The dark orange is the date range for the Laacher See, and the light orange the time before the drastic cooling.
There are several sulphate spikes, including two large ones. There are in fact 8 detectable spikes in the light orange zone. The very strong spike in the dark orange band is dated to 12,980 BP and would seem a very good candidate. However, it is also seen in Antarctica and therefore appears to be a tropical eruption. The same is true for the strongest spike, at 12,871 BP, which is also too late for the Laacher See. Five eruptions are in principle possible counterparts of the Laacher See, including the one we just ruled out. The other four have a very weak sulphate signal. Three of the four have a weak signal in Antarctica. The final one is dated to 12,985 BP (note: add 50 years for B2k dates), and is the most likely counterpart of the Laacher See. It is not impressive, but it is there.
Now all we need to do find some tephra in this ice core layer.
But this was too easy. Models for the eruption predict a larger sulphate deposit than found for this eruption. That is true even for low estimates for the total sulphate emissions. The large spike at 12980 BP would fit, and in fact the models suggest some of the sulphate will reach Antarctica, 5 times weaker than the signal in Greenland. This is about what is seen. So this suggests that the large spike in the dark orange band may be it, after all. Who said science was decisive? Any of the five could still qualify!
The large spike does in fact have some tephra in it. I have already mention it: it is the one with Hekla-like ash. This seems to rule out the Laacher See. However, there is an indication that the sulphate signal was a double peak. In that case it could be a combination of the Laacher See and another (Icelandic?) volcano. We may never know.
It is notable that there are four significant eruptions in a 110-year period, two of which were rather large (Tambora size). Both the large ones were probably in the northern hemisphere although not as far north as Iceland. The largest one was 25 years before the onset of the Younger Dryas. It was an impressive cluster of eruptions!
Abbott et al discuss whether this cluster could be the cause of the Younger Dryas. They decide against this, because of the 25 year delay. But could it have provided a push? Remember the Greenland Viking colonies? Their demise coincided with significant eruptions. It turns out that sea strait west of Greenland can cool dramatically after an eruption. Whether this pushed the settlements over the edge is something we may never know. Did the cluster make the warm climate of the time unstable, and thereby sowed the seed for the catastrophe of the Younger Dryas? The authors says that it is unlikely to be the sole cause, but that it may have acted as a trigger.
We have not found a single culprit. The consistency in time scales between the Dansgaard-Oeschger events and the Younger Dryas suggest ocean circulation is a major part of the puzzle. It may explain especially how it ended. A series of volcanoes, including the Laacher See may have provided conditions in which the Younger Dryas could erupt on an unsuspecting world. The 25 years delay makes this far from certain. But if it did, the Laacher See is co-responsible for one of the largest environmental crimes of the ages.
And it may tell us something about the fate of the Viking settlements. But in the end, catastrophes do not have single causes. Triggers only work if conditions are set.
The fault, dear Brutus, is not in our stars, but in ourselves
Albert, May 2022