During the last Ice Age, ice reached its maximum extension 20,000 years ago. Deglaciation started almost immediately and progressed rapidly. Sometimes ice meltwater would pile up behind an ice dam and when the dam collapsed a huge flow would follow. One such great flood occurred in the American northwest 13,500 years ago when an ice dam holding back about 2000 cubic kilometres of ice meltwater (Lake Missoula) collapsed. A huge mass of muddy water and debris rushed across the area into the Columbia River, cutting broad channels called coulees and forming the so-called Channelled Scabland … As a result of the flood that formed the Scabland, the sea-level rose very rapidly, from minus 100 to minus 80 metres [vis-à-vis today’s level]. By 12,000 years ago more than 50 per cent of the ice had returned to the ocean, and the sea-level had risen to minus 60 metres. At that point other giant floods occurred, down the Mississippi River valley into the Gulf of Mexico and down the Siberian river valleys into the Arctic Ocean. The Mississippi flood carried pebbles, which are now confined to the upper reaches of the Missouri-Mississippi system, all the way down to the delta. Sea-level rose very rapidly from minus 60 metres to minus 40 metres.34
The key phrase that caught my attention when I first read this passage was ‘ice dam’. It was very simple, and yet it explained so much. Averaged out over 10,000 years it was true that the total global sea-level rise of 120 metres at the end of the last Ice Age only amounted to a little more than a metre a century. But what Emiliani was now suggesting was the intriguing possibility that enormous quantities of the glacial meltwater could have been detained for thousands of years behind ice dams on continental Europe and continental North America – and then released into the open ocean all at once.
The ice-caps that formerly covered these areas were up to 4 kilometres thick, as we’ve seen, and larger than present-day Antarctica in both cases.35 Emiliani reminds us how:
The weight of the ice on the land surface below created bowl-shaped depressions about 1 km deep. Heat from the interior of the earth was trapped under the ice sheets, the bottom ice melted, and great freshwater lakes formed. Twice in North America and western Siberia these lakes busted through the ice margins and created huge floods. Sea-level rose abruptly around 13,000 years ago and again 11,000 years ago and then more slowly as the residual ice continued melting. Some have hypothesized that these prehistoric floods generated the flood legends common to many civilizations.36
Between 8900 and 8200 years ago, the Laurentide ice-sheet disintegrated in the Hudson Bay, facilitating catastrophic drainage of the massive Agassiz/Ojibway glacial lakes into the Labrador Sea. Based on Barber et al. (1999).
Professor Shaw’s abrupt steps
John Shaw, Professor of Earth Sciences at the University of Alberta, is one of the world’s leading experts on the last Ice Age and on its catastrophic meltdown. The author of an impressive list of peer-reviewed scientific papers, his research is at the forefront of inquiry in this field and has focused on the reasons for the superfloods. This is the graphic account that he gave us:
The big ice-sheets that covered Canada, most of Scandinavia and much of northern Russia – instead of them being pure ice and rock – it seems that at a late stage there was rock at the bottom and then a sub-glacial lake or reservoir of water, then the ice. And it’s possible that when warming occurred, the top of the ice started to melt, and the ablation zone and the sub-glacial water got bigger and bigger and bigger. And yet for good reason the ice-sheet seals around the edges. And then one time the big system on top connects – it’s a little bit like a toilet bowl, you sort of open the valve and the water comes surging through.
Graph of sea-level in the Caribbean against time since the LGM showing three abrupt steps around 14,000, 11,000 and 8000 years ago. Based on Blanchon and Shaw (1994).
Graph of rate of sea-level rise against time since the LGM. Based on Blanchon and Shaw (1994).
[In Canada on one occasion] the water literally came spewing out all over, except to the east of the Hudson Strait, because there was a big ice barrier there. So it came out southwards and through the St Lawrence, through the finger lakes, down through the Red river, South Winnipeg and the Winnipeg Lakes, and out through parts of Saskatchewan and out over the Milk river – which is the continental divide south of Alberta. The Milk river water flowed north to the Arctic, to the east to Hudson Bay, south to the Gulf of Mexico. And a huge amount of water went north into the Arctic Ocean. So you were suddenly introducing a vast amount of water to the oceans. And the duration of the flows was probably measured in weeks. And the kind of flow that we’re talking about, just for a small filament in Alberta, would have been 10 million cubic metres per second – that would drain Lake Ontario in about four days. And sea-level would have risen instantly, and somewhere in the region of 10 metres. This is about 15,000 years ago, when there were people living in many places. And the sea-level would have suddenly risen, and if you had lived by the sea-shore collecting jellyfish or something like that, and your house was suddenly underwater, you’d notice it. I imagine that it had quite an impression on the oral tradition and myths.
So the big event came from under the ice about 15,000 years ago. And then about 11,000 years ago there was a big lake in the southern part of the ice-sheet called Lake Agassiz that covered a big part of Canada. There was an equally big lake called the Baltic Ice Lake in Scandinavia. And then recent evidence suggests that there were big lakes across northern Asia and the north of the Soviet Union. These lakes were dammed by ice and tended to drain very suddenly. And as a result you get a similar effect, with a sudden rise in sea-level. Then last of all, about 8,000 years ago, there was the last lake in North America associated with the Laurentide ice-sheet which is called Lake Ojibway, and it lay just south of the Hudson Strait. And that lake drained catastrophically.
So originally it was thought that the rise of sea-level was steady at the end of the Ice Age, but now we are able to see that it rose abruptly in steps.37
Floods, volcanoes, earthquakes
Professor Shaw’s ‘abrupt steps’ were, arguably, the most traumatic experiences of global cataclysm that our species has ever undergone. To those alive then, the end of the last Ice Age with its sudden global floods must have seemed like the end of the world. Continental plates were shifting upwards relieved of the weight of the ice they’d supported for 100,000 years. Huge earthquakes and outbreaks of volcanism accompanied this extensive crustal rebalancing. The earth would have rung like a bell with tremendous sounds and vibrations. The sky would have been heavy with volcanic dust and black, bituminous rain. And at the same time the oceans were remorselessly, apparently unstoppably, rising. One of the geo-climatological mysteries of the last Ice Age is that the period of the meltdown – roughly from 17,000 to 7000 years ago – was also a period of dramatically enhanced volcanic activity. A paper published in Nature in October 1997 draws particular attention to what at first sight seems like a bizarre correlation between the rate of global sea-level change and the frequency of explosive volcanism in the Mediterranean area – with a distinct episode of enhanced volcanic activity registered in the geological and palaeo-climatological records between 17,000 and 6000 years ago:38
In areas where active volcanism and glaciation coincide, the correlation between the events can be explained by the effect of changing ice volumes on crustal stress. In contrast the effect of ice-sheet volume changes on unglaciated volcanic areas remains problematical. Several authors have proposed that meltwater loading and unloading could influence volcanic activity at sites distant from areas of ice-accumulation through the global redistribution of water, although this hypothesis has never been tested.39
The international team of scholars behind the Nature article counted tephra layers in deep-sea cores from the bottom of the Mediterranean (tephra is a general term for solid matter ejected during volcanic eruptions) and conclude that:
The frequency of tephra-producing events and, by proxy, notable explosive eruptions at Mediterranean volcanoes, can be re
lated to rapid variations in sea-level change. In particular we draw attention to the quiescent phase centred at 22,000 years ago and corresponding to the last low sea-level stand, and to the most intense period of tephra layer formation between 15,000 and 8000 years ago which accompanied the very rapid rise in post-glacial sea-levels.40
The authors think that ‘the existence of a single causal link between the rate of sea-level change and the level of explosive activity is unlikely’ and point out that ‘the unique response of individual volcanoes to large changes in sea-levels requires detailed study of each eruption record’.41 Where this has been done, however, ‘The level of explosive eruptions is seen to fall to a marked low between 22,000 years ago and 15,000 years ago, coincident with the last low sea-level stand.’42
I find it intriguing that the end of a 7000-year period of volcanic quiescence 15,000 years ago, and the beginning of the period of violent eruptions, both overlap with the first of John Shaw’s global superfloods; likewise the end of the period of enhanced volcanic activity around 8000 years ago follows Shaw’s third and last superflood.
Addressing this point, the scientists writing in Nature argue for broad-scale influences operating, for example, through
stress changes in continental margins and at island arcs. These may promote the ascent of fresh batches of magma into volcanoes, while increased levels of regional seismicity related to load distribution may play a role in destabilizing already weakened volcanoes.
On a global scale the number of volcanoes susceptible to the above-mentioned effects is large. Current spatial distributions of active volcanoes show that 57 per cent form islands or occupy coastal sites while a further 38 per cent are located within 250 kilometres from a coastline. Assuming a similar distribution for around 1500 volcanoes active during [the last Ice Age], then 1400 are likely to have been subject to the more direct effects of rapid sea-level change … Furthermore, the rapidity of these sea-level changes, and consequently their potential to trigger responses in active volcanic structures, are only now becoming apparent.43
Despite its authors’ caution about identifying a single cause, the evidence set out in the Nature paper does suggest that the earth’s own isostatic rebalancing process, sparked off by the sudden meltdown of the ice-sheets and rapidly rising sea-levels at the end of the last Ice Age, must have been what awakened the volcanoes. The implication is that isostatic adjustment does not always proceed at a constant, steady rate – otherwise volcanism would presumably be constant as well – but must at times involve large, rapid shifts transmitting shock-waves through the earth’s crust powerful enough to set the volcanoes raging around the globe.
It is precisely a shift of this speed and magnitude that Koudriavtsev envisages with his hypothesized ‘overnight’ collapse of the Celtic Shelf on the forebulge of the Fennoscandian ice-sheet 11,600 years ago. Moreover, researchers have found evidence that the meltdown of the same ice-sheet also unleashed tremendous forces during other periods of rapid worldwide flooding. At the time of Shaw’s third great flood around 8000 years ago, for example, the stresses and earthquakes became so severe that immense waves were formed in the ground. One of these, in northern Sweden, is 150 kilometres long and 10 metres high and has been described as a ‘rock tsunami’44 that can only have been caused by ‘earthquakes of unbelievable magnitude’.45
Descent of hell
Snaking across a bleak landscape, Sweden’s Parvie (‘wave in the ground’) as it is known locally, is a remarkable and somewhat disturbing feature, exactly resembling a three-storey-high tsunami made of solid rock caught forever in freeze-frame as it rears up just before breaking. The most remarkable – and disturbing – thing about it, however, is that this part of northern Sweden is a zone of extremely low seismicity and stands on what geologists define as a ‘stable continental region’ (SCR) of the tectonic plate.46 There should be no reason for catastrophic earthquakes ever to happen in an SCR. Yet the evidence unambiguously demonstrates that a catastrophic earthquake – indeed ‘the largest earthquake ever known within the stable continental regions’47 – did throw up the Parvie:
Studies over the last two decades show that it formed suddenly by earthquake faulting in the late glacial to early postglacial times of the great Fennoscandian ice sheet (approximately 8000 to 8500 years ago), suggesting a genetic relationship between the two.48
The precise nature of this relationship and the true magnitude of ‘post-glacial faults’ (PGFs) such as the Parvie have been studied by Ronald Arvidsson of the Seismological Department of Uppsala University. He has shown that such faults – of which there are a whole series in northern Sweden – frequently cut as far as 40 kilometres deep into the earth’s crust. All were caused by different gigantic earthquakes and all these earthquakes occurred within the same thousand-year period between 9000 and 8000 years ago.49
Arvidsson’s widely agreed estimate is that the Parvie quake measured 8.2 on the Richter scale.50 Another scholar, Arch C. Johnston of the Centre for Earthquake Research at the University of Memphis, points out that quakes of this magnitude only occur today along the edges of tectonic plates. The force that formed the Parvie ground-wave must, therefore, have been enormous:
The Fennoscandian PGF’s are … a remarkable consequence of rapid crustal unloading as the ice-sheets of the last Ice Age melted. The Parvie and other PGF’s … represent the faults of induced earthquakes, events that would not have happened without externally-imposed … conditions.51
Johnston then goes on to note that, although ‘induced seismicity’ is known today,
the post-glacial earthquakes are easily the largest known examples of this class. Surface quarrying can generate earthquakes of 2 to 4 [on the Richter scale];52 deep mining and deep-well waste disposal 5 to 6 events; and large hydro-reservoirs mid 6 events. Excluding PGF’s there are no earthquakes exceeding 7 confidently considered induced. The earthquake magnitude seems to scale with the agent of change of crustal stresses: great ice-sheets can induce great earthquakes.53
Now a characteristic of the Richter scale, not widely understood by those who live outside earthquake zones, is that it is calibrated so that each increase of one unit represents a tenfold increase in the magnitude of the quake.54 So a 2 is ten times bigger than a 1, a 3 is ten times bigger than a 2, a 4 is 10 times bigger than a 3, and so on. The earthquake that hit Kobe in Japan on 17 January 1995, killing more than 5000 people in twenty seconds, measured 7.2.55 With a Richter scale value of 8.2, the Parvie quake was ten times bigger than Kobe. The largest earthquakes ever recorded on the scale – rare events in subduction zones under oceans or between continental plates – have not exceeded the value of 9.56
The clear implication of Arvidsson’s and Johnston’s research, therefore, is that crustal rebound and isostatic rebalancing did at times take place very rapidly as the ice-caps melted down into cascading floods – rapidly enough to trigger extremely violent earthquakes and sudden massive faulting (penetrating to hitherto unheard-of depths of 40 kilometres and radiating laterally for up to 160 kilometres).57 Writing up his findings in Science magazine, Arvidsson concludes:
I interpret the earthquakes as signs of a progressive rapid rise of the land from the centre of postglacial rebound … to the outer reaches of the ice-sheet … More than 9000 years ago a nearly isostatic equilibrium was reached due to the depression of the lithosphere by the ice. After a quick removal of the ice-sheet a non-isostatic condition caused compressional stresses within the crust which triggered the earthquakes.58
Since the Parvie is only one of many giant post-glacial faults associated with the collapse of the Fennoscandian ice-sheet, what Arvidsson is really talking about – I think – is the descent of hell in northern Europe for a reign of 1000 years centred on 8000 years ago. As we follow his evidence, we must envisage extraordinary scenes of geological turmoil in which continuous deep tremors vibrate all the way through the Baltic Shield crust and the earth repeatedly roils, fractures, rears up and collapses – seemingl
y about to tear itself apart … While this is happening the ancient ice-cap over Fennoscandia is in a state of runaway meltdown, close now to the point of total collapse, and huge chunks of decaying ice the size of islands are falling into the sea, generating cataclysmic displacement waves. The ice-cap over North America is behaving in much the same way …
And let’s not forget that the earth by this time – 8000 years ago – has already suffered the consequences of 7000 years of intense volcanism, 7000 years of rising sea-levels and sudden and unpredictable marine floods, 7000 years of continental shelves, land-bridges and islands vanishing beneath the waves, and 7000 years of spectacular climatic instability. Indeed, the palaeo-climatological record testifies to all of the following – and much more – between 15,000 and 8000 years ago: cold oceans, high winds, mountains of dust in the atmosphere59 and wildly unpredictable temperature shifts.60
To give an example of the latter, Romuald Schild of the Polish Academy of Sciences cites an abrupt warming that took place in the northern Atlantic at around 12,700 years ago, stopped and equally abruptly went into reverse 10,800 years ago – when there was a sudden 800-year plunge to almost full glacial temperatures – then turned again to another episode of abrupt warming about 10,000 years ago.61 Robert Schoch reports that the bulk of the first warming ‘approximately 27 degrees Farenheit, a massive increase’ – occurred after 11,700 years ago:
Remarkably, the ice-core data suggests that half of the temperature change, in the neighbourhood of 14 degrees Farenheit, occurred in less than 15 years centring around 9645 BC. That’s a bigger temperature increase, and faster, than the scariest doomsday scenario about global warming in the twenty-first century.62
Underworld: The Mysterious Origins of Civilization Page 10