Glacial limits and ice streams in Wales

 Map of glacial limits for the Devensian in Wales, assembled by the Geological Survey from many sources, and further modified since.  The map also shows the contacts between Welsh ice and Irish Sea ice,  ice drainage channels, major glaciers and ice streams, and some features on the ground.

The arrows for Irish Sea Glacier movements are very generalised.  Some of them refer to the Devensian glacial episode, and some to the Anglian.  Almost all of the limits shown on this map are subject to further debate and correction.  But we are getting there.....

The Devensian ice limit in Pembrokeshire is mine  -- other published lines are demonstrably wrong, since they do not accord with the field evidence.

Click to enlarge.

Big Bang on Ellesmere Island

I posted this image some time ago, but the more I look at it the more extraordinary it seems.  It is indeed the glaciological equivalent of a Big Bang -- with seven glaciers flowing into the same small section of a deep glacial trough and ending up in a gigantic corporate collision.  This is seriously deranged behaviour -- how could it have happened?  Isn't nature normally rather more organized than this?  Well, yes it is -- so something very strange must have happened here.  All I can think of is that the big trough into which these glaciers are flowing is a very old relic of a previous glaciation, in which a major outlet glacier must have carried ice from an ice sheet or large ice cap towards the coast.  For some reason that supply must have been cut off, allowing glaciers from smaller ice caps to take possession of the trough.  But why are the glaciers flowing together like this, instead of establishing a common exit route?  I'm really at a loss to explain it -- and maybe I might get a clearer idea if I was to zoom out and take a look at the regional -- rather than the local - landscape.......

In the meantime, just look at the extraordinary compression structures on all of the glacier snouts. Click to enlarge the image.  There is also a lot of melting going on here.  Where does the meltwater go to?  Answers on a postcard please.....

Iceberg Breeding Ground -- Gaasefjord

This is an extraordinary "fluke" photo taken by satellite above Gaasefjord, one of the fjords running into Scoresby Sund, East Greenland.  Sorry if the photo and the caption appear pornographic.........

These are two tidewater glaciers ejecting a constant supply of icebergs and brash ice into the fjord.  We are probably looking at a six-month winter supply of glacier ice here which has simply piled up because it could not escape -- the photo was probably taken in May or June just as the sea ice was breaking up and "releasing" the accumulated iceberg debris so that it could be more easily dispersed into the fjord and onwards towards the outer coast. 

You don't often see a dense accumulation of iceberg debris blocking up a fjord like this -- by comparison, the icebergs (and brash ice) falling off the glacier snout into the sea in the summer are all quite rapidly dispersed by tides, winds and currents.

Peary Land glacier images

Ice Sheet edge in Peary land, north Greenland.  This might give a reasonable impression of what the ice sheet edge might have looked like in Somerset or Wiltshire during the Anglian glaciation.  This is a summer photo -- the land is snow-free for maybe 3 or 4 months in the summer.  Nevertheless, there is very thick permafrost.

A winter photo of an outlet glacier piedmont in Peary Land.  It's almost too perfect to be true!!  The ice in this glacier is very cold -- this would be classified as a "polar glacier" in which there is very little trace of meltwater action either in or beneath the glacier.

Erratic clusters and alignments

The other day I came across a very interesting study of glacial erratic distributions in Washington State, in the NW of the USA. Researchers have plotted the locations and types of thousands of erratic boulders on the flank of Rattlesnake Mountain.  The erratics apparently became visible for the first time following very extensive forest fires which effectively cleaned off the vegetation cover, leaving a bare land surface.  The interesting thing is that few of these erratics have been directly dumped by glacier ice coming from the Cordilleran Ice Sheet.  Instead, they have been carried and emplaced by floating and grounded icebergs in periodic paleofloods and temporary lakes -- with water up to 800 feet deep at times.

The erratics appear to be of many different ages -- suggesting that many have been picked up, carried and dumped on several different occasions, each one coinciding with a new glacial episode.

There are many single erratics, erratic clusters and erratic trails -- but in this case the trails appear to be perpendicular or transverse to the main direction of iceberg movement or water flow. Erratics do indeed behave in a myriad of erratic ways......



Extract:

The rocks were left after ice dams holding back a huge lake near Missoula, Mont., broke, re-formed and broke again from 1 million to 2 million years ago to as recently as 13,000 years ago. The floodwaters backed up at the downstream end of the Pasco Basin behind Wallula Gap, a narrow ridge opening through which the Columbia River flows today. The lake lapped the gentle slope of Rattlesnake Mountain, northwest of the gap and part of the Hanford Reach National Monument in south-central Washington Rattlesnake Mountain was the highest peak protruding from the 800-foot-deep temporary body of water, dubbed Lake Lewis. The rocks and boulders, so-called erratics, grounded as the waters of Lake Lewis receded after a few days like a slowly draining bathtub.
Bjornstad led a team that surveyed and analyzed the mostly-granite-strewn debris fields over 15 square miles of Rattlesnake Mountain. He said that the ice-rafted debris left deposits of three types: widely scattered rocks and boulders, distinct clusters and "bergmounds" – low, cone-shaped clumps of erratics that, like a moraine left by glaciers, alter the topography. Bjornstad's group discovered rafted rocks as long as 14 feet.
The erratics were concentrated along northeast-running gullies. Bjornstad suggests that the speed of the flowing water varied as it crossed an uneven surface, and that may have created eddies that forced an ice jam in the deeper, quieter waters at the back of these gullies. The erratics and bergmounds decreased as the surveyors worked up the mountainside. Bjornstad attributed the lower number to smaller successive floods.
Bjornstad and his colleagues found that most of the erratics were rounded, showing the effects of weathering and suggesting that they were carried in by older Ice Age floods.

------------------------

 Sources:
https://gsa.confex.com/gsa/viewHandout.cgi?uploadid=9
http://www.sciencedaily.com/releases/2003/11/031104064500.htm

British Glacial Limits

 Proposed glacial limits for the Anglian and Devensian Glaciations in Southern Britain 

(BSJ, Aug 2012)

Here we are.  This is my latest attempt at portraying the British Glacial Limits for the Devensian and Anglian glaciations -- along the southern margins of the British and Irish Ice Sheet.  I think it accords pretty closely with the evidence on the ground and with the glacial modelling which I have covered quite extensively on this blog over the past couple of years.  Current dating puts the Anglian at around 450,000 years ago, and the Devensian at about 20,000 years ago.  There may well have been another glacial episode between these two -- currently referred to as either the Saalian or Wolstonian Glaciation.  Many authors have attempted to map the limit of that glacial episode -- broadly, it seems to have been more extensive than the Devensian glaciation, and less extensive than the Anglian.  However, the line drawn by Gibbard and Clark is so unsatisfactory, in so many ways, that I have left it off my reconstruction above.  Let's just say for the moment that the events -- and the deposits -- of that episode are complicating factors, and that they will one day get sorted out........

Let's concentrate on South Wales and Southern England.  Note that on the map I have shown the Anglian and Devensian ice edges in more or less the same position on the north coasts of Devon and Cornwall, assuming that the cliff barrier was sufficient to prevent any great ice incursion inland during either of the glacial episodes.  In both glaciations the Irish Sea Glacier affected the Bristol Channel; in the earlier one the ice was so powerful that it pushed all the way eastwards into Somerset and maybe into Wiltshire, deeply inundating the whole of Pembrokeshire in the process; but in the Devensian the ice was more limited in extent, affecting Carmarthen Bay but maybe failing to penetrate further east than the Gower Peninsula.

Map showing the proposed extent of the Anglian Glaciation in Southern Britain, with approximate flowlines.  On this map the small ice caps and perennially snow-covered terrain of SW England are shown as being incorporated into the glaciated area.  In reality the margin between Irish Sea ice and locally generated ice caps would have been very difficult to discern.  The ice caps over Exmoor, Bodmin Moor, Dartmoor, the Blackdown Hills, the Mendips and the Cotswolds are shown with the dotted symbol and identified by letters.

One or two points relating to the Anglian.  The ice direction arrows are now pretty reliable, having been thoroughly established over many years by reference to striae, erratic transport, and glaciological modelling.  That having been said, there must have been a very complex contact zone in the South Wales coastlands between Welsh ice streaming southwards from the Welsh Ice cap and Irish Sea ice pressing in with great force from the north-west and west. The evidence for pushing the glacier into Wiltshire?  Well, glacial deposits on the Somerset coast, into the Somerset Levels and on the flanks of the Mendips, and glacial erratics on Salisbury Plain.  This very extensive ice cover is also needed to provide a satisfactory explanation for the giant erratics (apparently ice rafted) on the English Channel coasts. It's all to do with isostatic depression. See many earlier blog entries on this.   If anybody doesn't like this reconstruction, please give me a better explanation of these large coastal erratics......

One feature which is crucial to my reconstruction is the difference between the western and eastern pro-glacial zones -- ie the areas beyond the Irish Sea Glacier edge.  I think that in the west, in the counties of Cornwall, Devon and Somerset, there was considerable snow and ice accumulation during each glacial episode, giving rise to many small local ice caps and perennial snowfields.  David Evans, Stephan Harrison and colleagues have given us a good examination of one of these -- the Dartmoor Ice Cap.  Further to the east, the air temperatures and ground temperatures were lower, and there was not so much precipitation -- so the landscape was essentially one affected by frozen ground or periglacial processes.  The boundary between these two zones was probably in Wiltshire and Dorset.

A reconstruction of the Celtic Sea Piedmont Glacier during the Devensian maximum.  The map excludes the complex contact zone between Welsh Ice and Irish Sea Ice in South and West Wales, and the ice cap / perennial snowfield zone of Devon and Cornwall.

One of the key features of my Devensian map is the ice margin in the outer part of the Bristol Channel.  I think that the outer edges of the Pembrokeshire Peninsula were glaciated in the Devensian, since we find fresh glacial deposits near Dale, at West Angle Bay, and on the island of Caldey.  On Caldey the ice was moving pretty well west to east.  Maybe the ice margin was a little further eastwards, in Carmarthen Bay.  There is still doubt about the events on Gower -- I would appreciate input from others on this.  So was there effectively an ice dam across the Bristol Channel?  As the ice started to melt, could there have been a great glacial lake here?  There might have been -- and a reexamination of all of the submarine materials in the channel might give us some clues on this.  The Devensian ice also reached the Scilly Isles -- as established by James Scourse and others, with the aid of radiocarbon dating evidence.  However, we have to explain the very similar altitudes of the ice marginal deposits on all of these coasts, from St David's Head to the Scillies -- and if you reconstruct the surface contours of the Irish Sea / Celtic Sea ice mass, the contours must run broadly parallel with the ice edge.  That means that the ice must have been moving predominantly from the NW towards the SE.  I part company with most other geomorphologists on this -- they want ice in the Celtic Sea to be moving from NE towards SW, and they also want an ice surface gradient which is so low that it seems to me to defy the laws of physics.

So there we are, boys and girls.  That's my current best explanation of the situation on the ground.  Now let's test it to destruction........

Where was the Somerset GBG Limit?

 A computer-generated model of the area affected by the British-Irish Ice Sheet -- one of many models coming from the glaciology team at Aberystwyth university.  Note that according to this model Stonehenge lies within the ice margin.  But local ice caps are not shown -- and the modellers have not shown the Dartmoor Ice Cap....

In the light of the Dartmoor evidence which was recently published -- showing pretty unequivocally that Dartmoor and the other uplands of SW England were glaciated during the Devensian or LGM (Last Glacial Maximum) -- we have to do some serious revision relating to ice limits.  Although the evidence for the Devensian is easier to pick up than the evidence for older glaciations (including the Anglian) things are complicated by the fact that ice edges have overlapped in quite complicated ways -- with the "greatest extent of ice" being of one age here, and of quite a different age over there.  Each glacial episode and each growth of the British-Irish Ice Sheet has its own dynamics ...... and that makes life complicated for glacial geomorphologists.

That having been said, one must try to make some sense of the situation in the south-western counties of England.  As I have pointed out many times on this blog, there is physical evidence on the ground for a glacial advance by the Irish Sea Glacier across the Somerset coast and into the Bristol area.  There is further evidence of glacial material in the Bath - Bathampton Down area, and as far as the Greylake No 2 Quarry shown on this map:

 Relief map showing main uplands and lowlands, rivers and key Quaternary sites in part of the SW quadrant of England. (Source: GCR tome on SW England)

 Further inland, the evidence is more equivocal, and we still do not have clear evidence that the ice reached  Street, Glastonbury, and the SE end of the Mendips.   The glacial deposits so far recorded in this area appear to be very old -- they are conventionally interpreted as Anglian in age.  On the other hand the Dartmoor evidence appears to be supported by the glacier modelling experiments done at Aberystwyth University, showing that (on the basis of glaciological principles) the Irish Sea Glacier COULD have covered the whole of the Somerset Lowlands and reached as far east as Stonehenge and Salisbury Plain.  So where is the ground evidence to support that contention?  Maybe a mottley collection of erratics at Stonehenge is the evidence we need -- but I quite accept that not everybody is inclined to accept that!

But there is a real dilemma here --  why are the glacial materials already described in the literature low down in the stratigraphic sequence, with no fresher deposits above them?  But it defies logic to say that Devensian ice affected the SW uplands of Devon and Cornwall and reached the Scilly Isles, but did not extend as far as Salisbury Plain.  Even if the ice sheet had an extremely low surface gradient when it pressed in from the W and NW, it must surely have pressed as far inland as Yeovil, Shepton Mallet and Frome?    The laws of physics must apply -- although sometimes a simple law needs to be replaced by something more complex if explanations are to make sense.....

I still think it's probable that the Stonehenge erratics are genuine GLACIAL erratics, transported initially in the Anglian glacial episode.  But I also think that in the Devensian the area shown in the lower map was affected by a very complex set of processes in which low temperatures, periglacial action, snowfields, icefields, small local ice caps and the glacier ice of the Irish Sea glacier all interacted with one another -- with an emphasis on landscape protection rather than alteration.  Prof David Evans and his colleagues have pointed out that the ice on Dartmoor must have been thin, very cold, and very slow-moving -- so that there are very few traces of either erosional or depositional features in the landscape of today.  Also, most of the "erratic" material carried by the ice was not very erratic at all, having been moved maybe no more than a few tens of metres from its source areas. 

Ice thicknesses on the Dartmoor Ice cap at the time of its greatest Devensian extent.  (After David Evans, Stephan Harrison and colleagues.)  Note that in some places the ice was c 180m thick.

To synthesise. I think that during the Devensian maximum the whole of this landscape was covered by perennial snow and ice. Maybe for short periods in the warmer summers, the snow blanket was broken up, allowing solifluxion and some redistribution of materials to occur. How long did this snow cover last for? Centuries or millennia -- it's difficult to say, without hard evidence. Maybe it waxed and waned several times. Beneath most of the snow blanket, the ground must have been permanently frozen or locked into permafrost, but maybe some of the best-known periglacial processes (frost cracking, polygon formation etc) were inhibited because ground temperatures were not low enough. In hollows and valleys there must have been thick accumulations of snow, and there must have been many other snow-banks against escarpments and hillsides. The landscape might have looked like this from above:

 

Winter landscape, West Greenland, well beyond the ice sheet edge.  Some of these snowbanks are perennial -- but most of them melt away every summer.  The distance from the left edge of the photo to the right edge is less than 2 km.

Over the following uplands there were probably small, thin ice caps such as those I have often illustrated on this blog: Bodmin Moor, Exmoor, Dartmoor, the Blackdown and Brendon Hills, the Quantocks, the Polden Hills and the Mendips. As for the other smaller hill masses and upland plateau areas (including Salisbury Plain) -- they too might have supported thin ice caps at the same time as the Dartmoor Ice Cap was in existence. I also think that a lobe of ice from the Irish Sea Glacier must have pressed in from the Bristol Channel and Bridgwater Bay -- pressing inland for maybe 40 km or more. The ice also reached the Scilly Isles and pressed against the coasts of Devon and Cornwall, sometimes unable to push inland much beyond the cliffline and in other places making small incursions into old estuaries and embayments of lower land.

So what we have here is an extremely messy situation.  If you had flown over or walked about in the area at the time, you would have been hard pressed to define the junctions between Irish Sea glacier ice, local ice created on the upland ice caps, and firn and snowfields which were quite independent and more liable to seasonal melting and regeneration.  I will shortly post a map showing where I think the ice edge might have been.....