Some of the ideas discussed in this blog are published in my book called "The Bluestone Enigma" -- available by post and through good bookshops everywhere. Bad bookshops might not have it....
To order, click

Friday, 11 June 2010

New dates shed light on old ice limits

It looks as if this map, published a few years ago, is not far wide of the mark -- although I'll reserve judgment on whether the eastern edge of the Irish Sea Glacier impinged on the coasts of Devon and Cornwall. i think that it must have done, in tune with the physics of ice. There are seven new cosmogenic isotope dates from Pembs currently in the publication pipeline. They show as follows:

3 tors on Carningli -- dates all over 100,000 years.
4 igneous rock outcrops in the Dinas - Fishguard - Strumble Head area -- dates between 33,500 and 76,800 years (the last one, near Pwll Arian on the Pen Caer Peninsula, is somewhat anomalous but is explicable by reference to a short-lived ice cover and very little erosion. There might have been slightly more erosion on the other 3 tested sites.)

I should like to see more detail on sampling points and local geomorphology, since some might say that an ice cover around 20,000 years ago, if it was active enough and prolonged enough to leave moulded surfaces and to refashion coastal tors) should leave fresh rock surfaces giving dates of 18,000 years or so. So there are still uncertainties.

However, the clear difference in rock exposures inside and outside a postulated ice limit gives dates on the one hand over 100,000 yrs and on the other around 34,000 years. Sounds reasonable to me -- and this is in tune with some of my earlier posts relating to ice-moulded surfaces and the Pont Ceunant Moraine.

So the ice limit for the Devensian Glaciation in Pembs, as shown on the above map, seems about right. That means we cannot have ice pressing well inland in Somerset (and even reaching Salisbury Plain) in the Devensian Glaciation. Glacier ice does not do contortions like that! That means that the glaciation that carried the bluestones must have been far earlier, when Pembrokeshire was completely covered by ice -- as suggested from other lines of evidence mentioned in earlier posts.

Many thanks to Danny McCarroll for this new info. To be published soon:

Exposure-age constraints on the extent, timing and rate of retreat of the last
Irish Sea ice stream
Danny McCarroll, John O. Stone, Colin K. Ballantyne, James D. Scourse, L. Keith Fifield, David J.A. Evans, John F. Hiemstra
Quaternary Science Reviews xxx (2010) 1-9

Thursday, 10 June 2010

When were those stones moved?

From the above (click to enlarge) we can see that there are several candidate glacial episodes during which the erratics (and the Stonehenge bluestones?) might have been transported by ice of one sort or another. Look at the right-hand curve on the top diagram -- you will see that the cold (glacial) stages are shaded, and coincide with Oxygen Isotope stages 2,4,6,8,10, and 12. The most likely episodes are the Wolstonian (c 200,000 years ago) and the Anglian (c 450,000 years ago). The latter was a BIG glacial episode, with much more glacier ice in the Northern hemisphere ice sheets than at any time since. Rohling and colleagues think that there is very good evidence that global sea-level at this time was around -140m. That means that in order to maintain a relative sea-level somewhat equivalent to that of today (ie with the shoreline more or less where it is today on the edge of the Channel Coast landmass), there must have been c 140m of glacio-isostatic depression in Southern England and along the Channel Coasts.

See Rohling et al

Magnitudes of sea-level lowstands of the past 500,000 years
E. J. Rohling, M. Fenton, F. J. Jorissen, P. Bertrand,
G. Ganssen & J. P. Caulet
NATURE, VOL 394, 9 JULY 1998

Isostatic loading and rebound in Southern England

Giant's Quoit, Porthleven, nr Helston, Cornwall

Following my recent posts, I have been chasing around the geological and geomorphological literature on this topic, with remarkably little success. All I can find is statements like "There has been isostatic rebound in Scotland, but not in Southern England, where there was no glaciation." That appears to be the fixed view. Then, in all of the comprehensive descriptions of the erratics of the Channel Coasts and the coasts of Devon and Cornwall, many authors (even the learned fellows who contributed to the Geological Conservation Review for SW England) say "One possibility is that the erratics were transported by floating ice -- but on the other hand sea-level would have been so low at the time that that becomes a considerable problem." Some authors seem to get themselves into very contorted positions in order to avoid mentioning the phrase "glacio-isostatic depression." OK -- come on, you guys. If you don't want full-blown glaciation of the Channel Coasts (as proposed by Kellaway and others) you cannot explain the emplacement of all these boulders without considerable crustal loading of the south-west quadrant of England. And how do you get this crustal loading without an ice load? Answers on a post-card please.......

Wednesday, 9 June 2010

Southern Britain c 400,000 years ago?

Above: Imagine this as being the scene out to the west of the Scilly Isles during the Anglian or Wolstonian glacial episode, with the edge of the Irish Sea Glacier calving into deep water. (Actually this is the edge of the Columbia Glacier in Alaska)

Below: Ice foot and chaotic pressure ridge made of blocks of glacier ice and sea ice driven onshore. Somewhere on the south coast of England at the peak of the Anglian or Wolstonian glacial episode. (Actually this is Cape Murchison, Robeson Channel, between Ellesmere Island and Greenland.)

At the time, both land and sea might have been c 140m below their present-day elevations.

More on the English Channel ice foot

The Celtic Sea "glacier model" made by Dr Alun Hubbard and colleagues

I am more and more convinced that there is so much erratic material scattered along the coasts of the South-West Peninsula and the English Channel that we have to think very seriously about the role played by floating ice and the ice foot in the evolution of the coastline. Many years ago I wrote (with my colleague David Sugden) a hefty paper called "Coastal Geomorphology of High Latitudes" (Progress in Geography, Vol 7, pp 54-132.) In it, we included a pretty comprehensive piece about the ice foot, and touched on other matters including iceberg drift and sediment and erratic transport by floating ice. Recommended reading, though I say so myself.......

OK -- for these coastal processes typical of high latitudes to operate in Southern England, we need a very severe glacial / periglacial climate, floating glacier ice (ie a calving ice edge) somewhere appropriate for iceberg and ice-floe drift under the influence of winds and currents, and also sufficient isostatic depression to bring the shoreline down to the approx level of the sea surface of the time. We must be talking here about the peak of a glacial episode, with sea-level (Wolstonian? Anglian?) around 140m below the present. A simple calculation tells us that we need approx 400m of ice over the adjacent land surface, to achieve the necessary loading, or maybe even thicker ice a little distance to the north, and then a proportional or elastic depression of areas beyond the ice edge. The map above gives us just about the right scenario.

WITHOUT ICE FILLING THE CELTIC SEA AND THE BRISTOL CHANNEL WE JUST CANNOT GET THE RIGHT CONDITIONS TO EXPLAIN THE MASS OF FAR-TRAVELLED ERRATICS THAT WE FIND AROUND THE COASTS OF DEVON, CORNWALL, AND THE ENGLISH CHANNEL. Following deglaciation, isostatic uplift and eustatic sea-level rise have brought sea and land up to their present positions, although there may have been various transgressions and regressions during this process of recovery, since we cannot expect everything to have been in phase.

So let's take the map as a working model -- it will need modification, but it gives us the following:

1. A mechanism for transporting "bluestones" from the west into the environs of Somerset and the Wiltshire Downs.

2. A floating ice edge to the west of the Scilly Isles, providing ideal conditions for the calving of icebergs and brash ice into deep water.

3. A source for erratics and other debris to be carried by floating ice broadly eastwards into the English Channel, driven by prevailing winds and tidal and other currents.

4. An explanation for the erratics and till deposits (and ice edge meltwater channels) on the west-facing coasts of Devon and Cornwall.

5.Conditions perfect for the operation of "ice foot processes" along the shorelines of the Channel Coasts.

Hey presto. All sorted.

The ice foot holds the key?

More amazing info from Ian West's site:

Medmerry, near Selsey: At very low tides after exceptional storms more may be revealed. In late Victorian times there was much interest in the discovery here of a two ton mass of Bognor rock with striae like those in glacial deposits.

In 1892 Clement Reid discussed an unusual exposure of erratics at the base of the Ipswichian raised beach deposit at Medmerry. The exposure resulted from a storm of October 24th in 1891.

Reid 1892:
"About a hundred of these pits were examined, and the conclusion seemed irresistible that they afforded clear evidence of the agency of floating ice. Drift-ice grounding on the ancient foreshore dropped its burden of erratics between the tide marks. Here they were pressed deeper and deeper into the clay, for the rise and fall of the tide at high-water piled ice upon any projecting rock, while at low water the rock was pressed down by the weight of the ice till it was flush with the general surface. Often, however, the still-projecting boulder would be firmly frozen into a new ice foot, or accumulated mass of pack-ice, and would then be gently lifted out of the hole at the rise of the spring tides. It is thus that I account for the occurrence of empty pits, for they seem to mark the former sites of blocks which may have shifted their position several times before finally coming to rest. Perhaps some of the basins were produced by the stranding, packing and revolving of masses of ice during a storm, but the general appearance of the section suggests tranquil water in a sheltered bay. No signs of furrows ploughed in the clay were observed, and the ice was probably entirely in the form of flat-bottomed ice-foot, which, at a spot like this, sheltered from the prevalent winds by the Isle of Wight [but note that the prevalent wind-direction then is not known], would ground gently and would tranquilly melt away without being driven violently into the shoals, as on a more exposed coast."

A bluestone monolith from Chichester Harbour

From Ian West's stimulating site:

An exfoliating mass of pink granitoid rock measuring 1.8m by 1.5m by 0.9m was found west of the hardway on the north side of Chichester Channel at Itchenor Ferry. A block of speckled grey rock, approximately 3.1m by 1.7m by 1.4m was seen on the foreshore of bare London Clay 0.4m west of Longmore Point (White, 1915). Incidently this boulder was apparently not visible in 1891. In May 1914 it was 22.9m distant (seaward presumably) from the face of the cliff (White, 1915). This was the largest erratic in the area covered by the Lymington and Portsmouth Geological Survey Memoir. It shows sign of rapid distintegration along gaping fissures, and at its eastern end lie four detached masses (ranging up to 1.37m in diameter), which when in position probably added another 0.45m to 0.6m to its total length (thus it was originally about 2.3m in length).

White (1915) submitted the rock for petrographic study. Dr H.H. Thomas contributed the following note on the basis of a petrographic examination:

"E 11220. The specimen submitted has been sliced and proves to be a much crushed Quartz diorite. The rock has no similarity with anything in the south of England and Wales, except perhaps with some of the Pre-Cambrian Quartz diorites of Pembrokeshire. Quartz-diorites are common in the Channel Island and the North of France, and these localities would seem to furnish a more likely source than Pembrokeshire, especially as other boulders have been found on the South Coast, which, with good reason, have been considered to be of extra-British origin" - H.H.T.

It is just appropriate to note at this point that the Stonehenge Bluestones, although not diorites, are dolerites supposedly from Mynydd Preseli in Pembrokeshire.

See also: Belderson, Kenyon and Wilson, 1973

White 1915

Reid 1892

Kellaway et al 1975

Tuesday, 8 June 2010

Sea ice and giant erratics

Giant's Rock at Porthleven -- weighing 50 tonnes. Origin unknown.....

Freshwater Gut, Croyde Bay, N Devon. The big boulder (I shall say this only once) is of garnetiferous hypersthene-bearing granulite of gneissose type. Origin unknown.......

Having taken a look at Neolithic sea-levels, what about Interglacial sea-levels in Southern England? This is an interesting question -- one which has been answered (to some degree) by Ian West:

West, Ian M. 2010. Sarsen Stones and Erratics of the Wessex Coast


"The English Channel glacier theory is interesting and stimulating but it is not strongly supported by firm evidence. It does not accord with the evidence of the flint gravels, there is no evidence for a "Glacial Lake Solent". The erratics were probably mostly transported by floating sea ice. There is clearly is a problem concerning dates of transport of the erratics and relative sea-levels. A phase of ice-floes, significantly earlier than the Eemian (Ipswichian) Interglacial accounts most satisfactorily for most features. The source of the erratics of the Hamphire-Sussex coastal plain seems to have been largely the south or southwest, the coasts of Brittany, the Cotentin Peninsula and the Channel Islands, but some more rocks of more distant origin seem to be present.

Some sarsens may have been transported by ice or flood action in the braided rivers of the periglacial environment. This may explain examples at Milford and other places away from the raised beach, unless these (low) area were still within reach of sea-ice. The higher level sarsens of Portland require further consideration; they are somewhat mysterious but there is a possible source area on the chalk downs northwest of Weymouth from which they could have been transported in some way. However, the deposit in which they occur is not sufficiently understood and the possibility of a high level raised beach on Portland has not been eliminated.

The erratics of the low-level Hampshire-Sussex coastal plain were transported to the area earlier than 128 thousand years BP at an unknown date. They were present near the surface during the Devensian. In the early Flandrian those on the present coastal plain and others under the sea to the south or in the channel of the Solent were also on dry land. The deeper ones were submerged round about about 8000 BP and large areas south of Hayling Island, for example, in post-Neolithic times. The erratics include sarsens (greywethers) particularly just east and southeast of the Solent, but the known examples are smaller than the stones of Stonehenge, and often weathered and crumbly to some extent. Dolerite, like the Stonehenge Bluestones, does not seem to have been recorded. There is no evidence that the sarsen stones were transported into the area by man and ice was the agent. Evidence has not yet been presented that any sarsens or other rock types from here were used in any way in connection with Stonehenge, but equally there is no evidence against. They were used in wall-construction in historic times.

Studies of the examples of erratics offshore could shed new light on the theories of ice-transport and date of origin. Unusual rock types might more clearly indicate the source areas. Any possible evidence of human (transport?)is clearly worth seeking."


This is interesting too:

Boulders, Salcombe Fishing Grounds, English Channel

Hunt (1880, 1881, 1883, 1885) found a considerable number of foreign blocks in the Salcombe fishing grounds, some 30 to 50 km south of the Devon coast. Of 40 blocks described, there is granite, microgranulite, serpentine, syenite, gabbro, diorite, basalt, "diabase" (dolerite), trachyte, gneiss, quartz grit, conglomerate, sandstone and chalk flints and other rock types. They are discussed further by Prestwich (1892). The serpentine is precisely like the Cornish varieties. Surprisingly the other igneous rocks could not with certainty be ascribed to the English or French coasts. The gneiss resembled Hebridean gneiss from Scotland.


The famous pink granite erratic at Saunton, lying BENEATH the sandrock and other deposits dated to the last (Ipswichian) interglacial. This gives us a clue to the age of these erratics. They have been here for a very long time......

OK -- the erratics may date from a number of different glacial episodes, but the consensus seems to be that they were transported and dumped either in the Wolstonian glacial episode (around 250,000 years ago) or in the Anglian glacial episode (around 400,000 years ago).

Most (but not all) of them are beneath 10m asl -- but as many have pointed out, nobody knows how many similar erratics there may be in SW England buried in soils and slope deposits. Absence of evidence is not evidence of absence. The 'concentration" around present sea-level may be more apparent than real, since on the coastline overlying deposits have been eroded away, leaving the erratics behind on ancient wave-cut platforms. This has led geomorphologists to expend a great deal of hot air on ice-floe transport and iceberg transport. I'm not very happy with that, since all the evidence points to LOW sea-levels at all times when glacier ice might have been calving icebergs carrying big stones into deep water. Whichever glacial episode we are talking about, and however much glacial ice there might have been in the Bristol Channel and in the Celtic Sea, I cannot envisage sufficient isostatic depression and rebound to have matched an early interglacial eustatic rise of sea-level similar to that of the Holocene.

No -- I still prefer to think of ancient glacial deposits in the South-West. and this brings us all the way back to the glacial transport of the Stonehenge bluestones.......

Severn Estuary -- where was the Neolithic shoreline?

Turner, Allen and Rippon (Archaeology in the Severn Estuary 11 (2000), l-12)

This is a very complicated business, because of the sheer extent of the "levels" all around the estuary. All are agreed that there has been a substantial transgression between the Mesolithic and the present day -- and there are peat beds, submerged forests and archaeological finds to prove it. But there have been short-lived regressions and transgressions as well -- partly due to the complex interplay of eustatic and isostatic factors. The whole scene is also rendered complex by the huge tidal range of up to 15m, and the ongoing process of sediment dumping and redistribution on these vast mud flats. So far as I can see, none of the authors studying the estuary has been able to draw a line on a map and say with conviction: "This is where the Neolithic shoreline (or mid-tide mark) was."

The Neolithic shoreline

This Landsat image gives us a clue as to what the coastline might have looked like during the Neolithic, with sea-level about 6m lower than it is today. Around most of the cliffed coasts the difference would have been negligible, except maybe for the emergence of a "fringe" of bouldery beaches along the base of many cliffs. In bays filled with sediments, LWMST would have been further offshore, and the inner parts of bays would have been densely wooded and probably boggy. The most dramatic changes in the configuration of the shoreline would have been in the inner parts of the Bristol Channel and the Severn Estuary.

Loughor Estuary at low tide. These gravelly and muddy flats would have been permanently exposed above HWM in the Neolithic. They would have been heavily wooded too, in sheltered locations.

Monday, 7 June 2010

Recent sea-level changes

This shows how sea-level has risen during recent times -- including Mesolithic, Neolithic and Bronze Age. Note that the dates are BP, not BC.

Post-glacial sea level rise, derived from data from many locations worldwide. There is still some doubt as to just how low sea-level dropped at the peak of the Last Glacial Episode. In the early stages of deglaciation, sea level rose at a rate of 1m per century -- but then came Meltwater pulse 1A, when sea level rose approximately 20 m over a 500 year period, about 14,200 years ago. That was a phenomenal rate, considering that we are talking here about the GLOBAL sea-level. The early dramatic rise coincided with the catastrophic collapse of the Laurentide Ice Sheet (North America) and the slightly less dramatic collapse of the Scandinavian Ice Sheet.

These raised beach terraces in NW Scotland were formed originally at a much lower altitude -- below present sea-level. But they have been lifted by post-glacial crustal rebound to c 30m above present sea level. This uplift has outstripped the eustatic rise in sea level.

Submerged forest (revealed at low tide) at Marros on Carmarthen Bay. As in all of the submerged forests around the West Wales coast, we can see fallen branches and tree trunks, peat beds, roots and stumps, and even accumulations of hazel nuts with the kernels still in place. They date from around 7,000 years ago, when Mesolithic hunters wandered these woods. The woodlands were later inundated by the rising sea-level, which here outstripped the isostatic rise of the land.

Sunday, 6 June 2010

On postglacial rebound and relative sea-level

I'm often asked where sea-level (ie relative sea-level) was in the Neolithic and the Bronze Age of Southern Britain. Well, we are still not entirely sure. Two factors come into play -- isostatic rebound following the melting of the "glacier ice load" over parts of Britain and Ireland, and eustatic sea-level rise consequent upon the return of vast amounts of meltwater to the oceans after c 20,000 years ago. So sea-level was rising, and so was the land -- and both were jerky rather than smooth. Both eustatic and isostatic rises were initially very fast, gradually tailing off to near zero. So shoreline sequences are often very difficult to interpret. The maps above show some of the complexity in Scandinavia. The highest shorelines are not necessarily in the areas of greatest rebound -- since many of those areas were covered with ice until later stages. There was a minimum rebound of over 275m near the head of the Gulf of Bothnia -- that translates into an ice load which was c 900m thick. In the same area the current rate of uplift is c 9mm per year.

But actually the rebound was much more than that, since sea-level at the time of initial isostatic recovery was around 120m lower than it is today. That means the TOTAL amount of recovery over the past 20,000 years has been at least 400m and more likely 500m. That equates to an ice load over the Gulf of Bothnia which was 1.5 km thick. That's a lot of ice.

Then things get REALLY complicated, because water that floods in to replace ice also has an isostatic effect, depressing the land surface over which it floods -- or at least slowing down the rate of ice-melt-induced uplift. There is also an elastic effect, which means that land surfaces up to 150 km beyond an ice edge can also be depressed by ice loading, and then affected by rebound when the glacial episode comes to an end. And there is a compensatory effect too, with UPLIFT beyond an ice edge caused by compensatory flow within the Earth's mantle when a glaciated area is depressed. When post-glacial uplift begins, this compensatory flow is reversed, so that areas beyond the ice edge may actually sink. Between these two zones there may be a hinge-line along which isostatic effects are zero. Add to that the complicating factors of rock type, fissures and faulting, and even tectonic movements that have nothing to do with glaciation, and confusion begins to set in. And by the way, vast loads of glacial and fluvio-glacial sediments transferred from an area of glacial erosion to an area of glacial deposition (for example, from the uplands of Britain into the Irish Sea and Celtic Sea) will also trigger isostatic effects.

In the old days, when I was a geomorphologist, I had to try, in places like Greenland, Antarctica and Iceland, to unravel the history of raised shorelines, ice edge positions, isostatic uplift and depression, and eustatic oscillations -- it was the closest I have ever been to four-dimensional puzzle solving!

Confused? So am I.......

Now to Britain. The map above shows that there is still slight uplift going on in the northern parts of the British Isles, but that there is sinking in the south. This sinking might be partly a compensatory isostatic effect, partly tectonic, and partly because of the massive sediment load carried by rivers into the southern North Sea and English Channel. The rate of uplift in the Highlands of Scotland is only about one third of that of the Gulf of Bothnia, since the British-Irish ice mass was that much smaller and since isostatic recovery is now more or less complete.

The highest post-glacial or Holocene shorelines in Scotland are nothing like as high as the shorelines I have studied in the Arctic and Antarctic. There are some at c 45m in south-west Scotland and Northern Ireland, although most of those identified have been beneath 30m. In West Wales there is no evidence of post-glacial shorelines above present sea-level; this means that the amount of isostatic recovery following the removal of the ice load has been less than the 120m or so of eustatic sea-level rise.

If we assume that isostatic uplift in South-west Britain is more or less complete, we would expect to find some evidence of prehistoric shoreline occupation BELOW present sea-level. This is just what we do find. During the Palaeolithic sea-level was more than 20m lower than it is today. During the Mesolithic c 7,000 years ago, sea-level was about 10m lower than today, and we find evidence of the Mesolithic sea-level rise in the submerged forests around the coasts of Pembrokeshire. In the Neolithic, sea-level was at c -6m, and in the Bronze Age it was at -4m. Around the time of Christ, it was about 1.5m lower than at present.

This all means that in many locations (for example, on low coastal forelands or in estuarine environments) HWMST would have been located out beyond the position of the present coastline. This is something that does need to be borne in mind by those who argue for the human transport of bluestones by land and sea. To my mind, the degree of difficulty would have been greatly increased -- with extensive boggy and heavily-wooded tracts to be negotiated in places where there is now sea.

Click on the maps to enlarge them.

Friday, 4 June 2010

And now for something completely different......

Fred Flintstone (or whoever it was) obviously didn't know his bluestones from his sarsens, but never mind.......

Sarsens map

This is an interesting map from Ullyott and Nash 2006 -- showing just how widespread sarsens are across the South and South-East of England. They are on many different rock-types and at many different altitudes.

Click on the map to enlarge it.