In the previous post Gijs described the surprising history of Europe’s most underestimated volcanic field. Much of the time the Eifel does nothing. On occasion it does a small eruption, just when you weren’t looking. And very rarely, it goes big, covering the surrounding areas in meter-thick ash.
Gijs described the area, the monogenetic volcanoes, and especially the Laacher See, the hole left by the largest eruption in Western Europe since before the ice age. The local deposits are tens of meters thick. It happened during a warm interlude in the ice age, in a newly forested area. Those forests across the region died, and took decades to recover.
Some are born great, some achieve greatness, and some have greatness erupted upon ’em.
What would a recurrence be like? (It wouldn’t be at precisely the same location because the volcanic field is monogenetic, i.e. every location erupts only once. It would be somewhere nearby.) Remember how even a small eruption in Iceland in 2010 caused transport chaos across Europe. It stopped planes from flying. A small ash cloud that would have had no impact in the stone age, became a disaster in the modern world. Nowadays, effects far away from an eruption can be more severe than those nearby, as was the case for Eyjafjallajökull. And the Laacher See was not a small eruption: this was a VEI-6 in the industrial heartland of Europe. The impact would be felt far and wide.
In a recurrence, the explosion will be heard across Europe. Imagine much of Western Europe in darkness for a day, pyroclastic flows reaching tens of kilometers from the site, and the city of Koblenz, 15 kilometers away, buried under 1.5 meters of ash. The airborne ash affects 14 separate countries. Rafts of ash block the Rhine: when the dam breaks, the Rhine delta floods and the ash rafts reach England. Transport comes to a halt for days to weeks. Six million people are significantly affected, from Bonn to Amsterdam. Health care is the biggest casualty of the eruption.
Of course the chance of a 1-in-100,000 year event happening next year is not very high. There are more likely events, such as the 1-in-500 year rainfall the region experienced last year, and the 1-in-100 year pandemic. But then, we weren’t prepared for those either, in spite of the certainty they would happen again.
This post is not about this. It is about time.
The deposits of the Laacher See have created a marker in the soil which is used to synchronize archaeology across Europe. It acts as a time signal. You can compare it to the cannon on Signal Hill which still fires every day at noon, so that ships can set their on-board clocks. (It is one of only two cannons in the world with a twitter account. It tweets ‘Bang’ once a day.) Of course nowadays this is for the tourists, not the ships which all can use GPS for their time signals. But volcanic soil markers are like a watch with only an hour hand and no numbers: you can see time progress but still don’t know what the actual time is. Thus it is with the Laacher See. This post is about dating the eruption with water.
But that is for later. The Laacher See eruption came just before the onset of the Younger Dryas, when Europe suffered a sudden and severe set-back in its ice-age recovery program. It was portentous: a sign of a disaster to come.
The Younger Dryas also has a dating problem, written in water. So let’s start with water.
The weight of water
We could not exist without water. A drink every now and then is essential to avoid dying by dehydration. But water is a strange substance with strange properties of physics which life has come to depend on. It can crawl up the inside of a stem – hence we have plants and trees. The solid phase floats – hence we have ice on the surface rather than on the bottom of the sea, providing a home for land-dwelling life on the ocean. Water is strong: you can walk on water – if you are an insect. But water also has some properties that are more detrimental to life. It makes the ground unstable. Even a thin layer of ice can cause heavy rocks to move.
For an inert substance, water is also amazingly explosive, as Hunga Tonga proved. The Laacher See started as a powerful hydrothermal explosion. In Europe’s snow catastrophe of 1979 (this was the era before global warming took off, now hard to remember) I saw houses buried to their roofs under snow, in an area that now gets very little snow.
That snow can also record our climate. That is part of our story of the Laacher See.
Volcano’s evil manners live in brass; their virtues we write in water.
Snow provides a historical temperature record. Obviously, this record is no more permanent than the snow itself, and so the record exists only where snow can survive the summer. One such place is the glacier of Greenland, which stores snow that fell more than a hundred thousand years ago. In Greenland the ice age still lives on.
But how does snow record the ancient temperatures?
Water comes in three different phases: solid, liquid and vapour. Each of these comes in many different forms. The solid phase forms ice but also snow. The liquid phase forms oceans, but also rain or drizzle. The vapour phase forms clouds, or mist, or fog. The UK version of English has 104 expressions for rain! (but only 10 expressions for falling snow). The Inuktitut language has 93 different words for sea ice. In the end, all these different phases, phrases and expressions describe a single molecule: H2O.
But not all H2O molecules are the same. Most of the oxygen in H2O is 16O, the common isotope made from 8 protons and 8 neutrons. But a small fraction is 18O, a slightly heavier form of oxygen which carries two extra neutrons. The H in H2O can also come in two forms: normal hydrogen consisting of just a proton, or heavy hydrogen (called “deuterium” or just ‘D’) which has an extra neutron. Combining all, there are 8 possible varieties of H2O, which carry between zero (for H2O) and four (for D218O) excess neutrons. The most common form after plain H2O is H218O which accounts for 0.204% of all our water.
The excess neutrons make the water heavier and more sluggish. Think of the neutrons as a company management team: more managers make things more stable but slow everything down.
The sluggishness of the heavier forms of water shows up well in the Greenland ice cores. The ice comes from water evaporating from the oceans, forming clouds, precipitating as snow, and over time being pressurized into ice. The problem is the first step, the evaporation. The lighter water (with 16O) will evaporate but the heavier water (with 18O) is less nimble and is more likely to stay behind. The snow that precipitates from the clouds over Greenland therefore has less 18O. It is a small difference, but measurable. Snow in Greenland is down by about 3% compared to the water it came from. ( 18O accounts for 1.98% of the snow, rather than the original 2.04% before evaporation.)
When it is warmer, the higher temperature overcomes the sluggishness and more 18O joins the clouds that make the snow and ice. When it is colder, more 18O stays tucked up in the warm ocean and refuses to get up and join the action. That difference shows up in the snow. Measure the amount of 18O in the precipitation, and it tells you the temperature. In this way, by measuring the amount of heavy water in the deep ice cores, we can tell how the climate has changed. And because the ice cap is so old, we can see the temperature variations over a period spanning more than 100 millennia.
Strictly speaking, this is the temperature of the sea and air where the water evaporated. And there is a complication: if the cloud produces rain or snow before reaching Greenland (as it will), that precipitation will remove more 18O (it drops out more easily, being heavier) and this increases the deficit. But it appears that the fraction of 18O in the ice remains a very good indicator of the average annual temperature. Just don’t ask to put actual temperatures on the scale.
The 18O abundance is presented as the difference with respect to the 18O abundance of average ocean water. The ‘average ocean’ is (surprisingly) an official standard. It is rather better defined than for instance the ‘olympic swimming pool’ you’ll find quoted in newspapers. (An olympic swimming pool is 50 by 25 meters, and at least 2 meters deep. Let’s call this one OSP. The ‘at least’ is the tricky part, as they may be deeper. It means that if you carefully measure out 1 OSP, and use it to fill an olympic swimming pool, it may not be full.)
The ‘average ocean’ is called the ‘Vienna Standard Mean Ocean Water’ (VSMOW).
If it is not immediately obvious why land-locked Vienna has ocean water, the reference water is maintained by the International Atomic Energy Agency which is based in that city. It was created from a mix of water collected from the surface of the ocean at locations across the globe. Reference samples are for sale from IAEA. Don’t rush to buy some: it costs an eye-watering (pun intended) 10 Euro per milliliter. VSMOW is used to define a standard for the abundances of isotopes of hydrogen and oxygen – and it is used for nothing else. Why the Atomic Energy Agency? The original goal was to be able define a baseline against which to measure pollution from atomic incidents.
The samples are completely purified: everything oceanic (such as salt and calcium) has been removed, leaving only the H2O and nothing else. You might think that water supplied by the Atomic Energy Agency would be more exciting. Drink it and you will be fine, severely disappointed, and a lot poorer.
I to the world am like a drop of water
That in the ocean seeks another drop
Scientists measure the fraction of 18O in an ice core, divide it by the fraction in VSMOW (making sure they measure it in exactly the same way as the ice core water), and subtract 1. As an example. current snow in Greenland has an 18O fraction of 0.0198, while VSMOW has 0.0204. Dividing the two gives 0.97. Subtracting 1 leaves -0.03. The minus sign show that the snow has a deficit. Converted to a percentage (multiply by 100) it becomes -3%. This deficit is called δ(18O). By tradition it is presented in permille (parts per thousand, ‰) rather than percent (parts per hundred, %), so the paper that reports the number will say that δ(18O)=-30‰. It will almost always be negative, as otherwise 18O (heavy water) would need to evaporate more easily than 16O (light water) which is difficult to achieve. But if you were to evaporate an entire sea, the residual water would have a positive δ(18O).
The number for δ(18O) is now used to obtain the temperature.
This is the part of science no one talks about – it is not glamorous and is unlikely to ever make the newspapers, but it is essential for converting field measurements into meaningful numbers.
Written in water: ice age’s end
The scientists take some ice from the core, at regular depths, melt it and purify it, measure the abundance of the 18O and D isotopes, and compare this to the official standard to calculate the deficit in the ice. During the depth of an ice age, that deficit becomes larger.
The figure shows the resulting plot of the 18O abundance in the Greenland ice core, from 130,000 years to 8,000 years ago. The vertical axis shows the deficit of 18O (the numbers are negative) in permille, i.e. -40 means -40‰, or -4%. I have taken the data from the recent re-analysis of Gkinis et al (2021): “A 120,000-year long climate record from a NW-Greenland deep ice core at ultra-high resolution.”, published in Science Data 8, 141 . The data is available from https://doi.org/10.1594/PANGAEA.925552
Now we are ready to investigate the ice age and it’s ending around the time of the Laacher See eruption. The temperature record starts with the previous ‘interstadial’, the brief warm phase between ice ages that happened 120,000 years ago. Interstadials typically last around 10,000 years. The interstadial ended with a sharp cooling which took place over several thousand years. Between 110,000 and 80,000 years ago the climate was chilly (‘frozen’ might be a better description but I’ll let it go). Between 80,000 and 60,000 years it became a lot colder (‘deeply frozen’). After that the world re-warmed a bit, but there were many intermittent, fast changes between the ‘frozen’ and ‘deeply frozen’ state. This period culminated in another very cold but brief phase 30 thousand years ago.
Never-resting time leads summer on
to hideous winter
The intermittent, fast changes involved a large number of fast warming spikes (’warm’ is relative), lasting typically a few hundred years. They would start with a sudden upward shift in 18O (a warming ocean) followed by a slow decrease back to the previous values of the deep freeze. The sudden warming would happen within decades. At times these temperature spikes were numerous, and other times they were few and far between.
By 25,000 years ago, a slow warming began to set in, as the ice age was drawing to a close. This warming continued with ups and downs, until 14,690 years ago (b2k means ‘before the year 2000’) when there was another sharp warming spike. Because there already had been some warming, and because this spike was larger than most, it caused temperatures to rise to values approaching those of an interstadial. The ice age had ended, the glaciers were melting, and the Bølling–Allerød interstadial had arrived. Finally, there was the promise of summer.
The Younger Dryas
And through this distemperature we see
the seasons alter: hoary-headed frosts
fall in the fresh lap of the creeping oak.
Except that it wasn’t. This summer was a false dawn. Temperatures slowly dropped back over the next 1500 years, becoming chilly, but perhaps feeling mild after so much cold. The forests that had returned to much of Europe survived the slow cooling, and even two brief phases of colder weather.
Disaster struck in 12,846 b2k. Across Europe, temperatures suddenly dropped, much faster than before. Within a century, and perhaps much faster, the continent was back in deep ice age conditions. The new-born forests died and tundra returned. This was Ice Age Resurrection. This was Frozen, that brilliant movie about a battle between cold and warm, fought by two sisters each with one foot in either camp, one with a frozen soul but a warm heart, and one a warm soul but a frozen heart.
As the trees died, a particular plant took advantage and became widespread. The small Alpine daisy is called ‘Mountain Avens’ or ‘Creeping Oak’. The official name is Dryas Octopetala (the latter meaning ‘8 petals’ – it sounds so much more impressive in sciency language). The re-emergence of this plant, where the real oak was replaced by this creeping oak, gave the cold snap its name: it became known as the Younger Dryas. (‘Younger’ because there had been two brief episodes of cooling already during the warmer weather, called the ‘Oldest Dryas’ and the ‘Older Dryas’. Room has been left for a ‘Youngest Dryas’.)
The cold snap of the Younger Dryas lasted 1200 years. It affected Europe, Asia and America. The southern hemisphere was not affected, at least not at the same time.
Why, what’s the matter,
That you have such a February face,
So full of frost, of storm and cloudiness?
After 1200 years, almost instantly, the warmth came back, and this time it remained. The holocene summer had begun.
After this ice age intermezzo, it is time to return to the Laacher Sea.
The Younger Dryas dramatically changed Europe. The forests disappeared, and tundra returned. The human populations which had expanded into the newly temperate regions left again for the south – or died.
But strangely, the deterioration in the climate in Europe came more than 100 years after the cooling in Greenland. The reason for this delay is opaque. One suggestion is that the sea ice took this long to grow, after which it pushed the storm tracks further south. There are other suggestions, but none are convincing. The delay is a mystery. And the mystery is closely related to the Laacher See eruption.
that one might read the book of fate,
and see the eruptions of the times
Counting time is a skill. In Mary Poppins, it is perfected by the chimney sweeps with their Step in Time, performed while covered in ash in a way that is strangely reminiscent of timing volcanic activity. Marie Poppins missed a trick here. But there are other ways to count distant time.
When the age is in, the time is out
The ice cores from Greenland show individual year layers which can be counted, but only for the past 2000 years or so. Further back, the layers merge into each other and we can derive approximate ages (meters of ice per century) but not individual years. The depth profiles are calibrated with known eruptions which have left deposits in the ice – which the Laacher See eruption did not do.
Tree rings also count time, and they can go a long way if we use several trees which died at different times. The trick is to match up overlapping parts of the profiles, for instance a sequence of 5 narrow rings seen in two trees may be coming from the same climate event. The uncertainty increases when we go back more than 8000 years, especially during cold periods when there weren’t many trees around.
Finally, lakes can also count, another time signal written in water. Sediment at the bottom of lakes in western Europe forms annual layers called varves. Layers with volcanic ash can be used to assign specific dates, and after that it is matter of recognizing and counting the varves. Thinner layers can be hard to distinguish, and this leaves a counting uncertainty of some 30 years.
Both the Laacher See eruption and the onset of the Younger Dryas in Europe are well seen in the lakes, and from this we know that the eruption pre-dated the Younger Dryas by some 180 years. The eruption caused an exceptionally thick varve in the local lakes, for obvious reasons, and this is easily recognized.
This is a plot of tree pollen in the Meerfelder Maar, another lake in the Eifel (much older than the Laacher See). The amount of tree pollen is measured in each of the different varves and plotted against time; time runs from right to left. The grey line shows the Laacher See eruption.
Before the eruption, tree pollen was plenty (mainly birch and pine trees). Immediately after the eruption, tree pollen completely disappeared for perhaps 20 years, and then slowly recovered. Clearly, the eruption had killed all the local trees (even quite some distance from the Laacher See: the two lakes are 30 kilometers apart!), and they needed time to regrow. Some 180 years later, the pollen suddenly plunged and now it remained low. This was the start of the Younger Dryas. Many of the trees that had regrown were now killed by the abrupt climate change: they froze to death.
The date of the onset of the climate catastrophe, as measured in this dataset, is 12,730 b2k. Remember that in Greenland the cold had come in 12,846 b2k.
The next plot, taken from Brauer et al. 1999, Quartenary International, 61, 1, shows the time lines from a lake in Europe and from a Greenland ice core. The Laacher See eruption is shown as ‘LST’: it left a clearly identifiable trace in the varves in the lake. The ice core has two volcanic sulphate peaks around 13,000 years ago, listed as T1 and T2, but neither of these coincides with the Laacher See. The plot again shows a clear time difference between the onset of the Younger Dryas in Greenland and in Europe.
Comparing the varve counts in the lake to the layers in the ice core indicates that the Laacher See erupted 10-15 years after the onset of cooling in Greenland. But when the North Atlantic froze, it seems Western Europe still maintained milder weather, with even a touch of warming summers, as if subtropical air found its way to Western Europe – until suddenly, 180 years later this stopped and temperatures finally plummeted across Europe. None of the climate models could explain this time lag. It was a mystery.
It was a mystery that has now been solved.
Where a volcano is dating an ice core, there is an obvious risk of an incompatibility – it is after all throwing hot at cold, like an ancient Frozen. In cases like this, a match maker is required. Frederik Reinig and collaborators stepped up to the challenge.
Their match making was published in Nature, 595, 66 (July 2021). They had looked for trees – not normally difficult in Germany (a country that loves its forests), but their trees were not in the land of the living. They looked for trees that had died in the Laacher See eruption. And with success: three remnants of trees were uncovered that had been buried in the tephra, and had succumbed to the eruption. The outermost wood on one of the trees showed that the annual growth in the final year had already started: the eruption had therefore occurred in spring or early summer.
Counting tree rings (not easy in carbonized wood) showed that the trees had been from 50 to over a 100 years old at the time of the eruption. Carbon dating was done on each annual ring (as far as possible). The time around the Laacher See eruption is a difficult one for carbon dating. The amount of C-14 declined a bit in the atmosphere at the time, in such a way that the measured carbon-14 age is almost constant over a period of about 100 real years. That has caused large uncertainties in C-14 dates for the Laacher See. By using the full range of tree rings, it was possible to make a plot of C-14 age versus tree-ring years before the Laacher See eruption. That plot was shifted in time to find the best fit to the known C-14 variations. This was repeated for each of the three trees.
This procedure gave a much more accurate age for the Laacher See eruption, and it was not the same as the age established from varve counting. The eruption had occurred in 13,056 ± 9 years b2k.
The new date made the Laacher See eruption older by about 130 years.
The varves show that the Younger Dryas started about 150 years after the eruption in the region, or around 12,857 b2k. This of course also is earlier than before, by about the same 130 years. The new age now fits the cooling in Greenland. It turns out that Europe hadn’t had an Indian summer after all. The cold weather had started here at the same time as elsewhere around the Atlantic. The mystery was solved: we had misread the clocks.
In the old dating, the Laacher See eruption happened close to the start of the Younger Dryas in Greenland. The new date puts it 150 years earlier. It leaves no doubt that the eruption could not have caused the dramatic cooling. It wasn’t there at the right time.
It is very difficult to terminate an ice age. The Younger Dryas shows how quickly it can come back. If a volcano had been to blame, that would have meant the same could happen nowadays. It seems it wasn’t. It is one less volcanic disaster to worry about.
Of course, we now have other climate clangers to concern us, with global heating and rising sea levels. But that is a problem we have made ourselves, and will have to solve ourselves. Volcanoes are neither the cause nor the solution.
The Laacher See VEI-6 eruption was the largest volcanic explosion in Western Europe for perhaps 100,000 years. A recurrence would not cause a climate catastrophe. But even without that danger, one wonders how well prepared Europe really is for such an event in its heartland. Germany (like the rest of the world) was prepared for neither the flood nor the plague in the last two years. Would we do better against fire? Remember that for global heating, we know exactly what is coming, and fairly well when. And still we aren’t prepared. We should be living in hope and planning in fear. But in reality we live in fear and plan in hope.
O, how shall summer’s honey breath hold out
Against the wreckful siege of battering days,
When rocks impregnable are not so stout,
Nor gates of steel so strong, but Time decays?
Volcanic eruptions create markers in time. What we do with those times is up to us.
So much for the Laacher See dating of an ice age event during the early days of Frozen. The story begs one question: if the Laacher See did not cause the Younger Dryas, and we have shown previously that the Hiawatha impact is also innocent, then what did cause the Return of the Ice Age? And just as mysterious but often ignored: why did it end so suddenly? That is another story.
Albert, with contributions from William (Bill) Shakespeare