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

Tuesday, 15 February 2011

A Glaciological Dilemma

Following on from my last post, this map (from Hambrey et al 2001) is important because it shows that at the time of the last glacial (Devensian) maximum in West Wales, the ice edge on the Pembrokeshire coast was approximately at the altitude of present sea-level.  (Let's forget about eustatic and isostatic variations for the moment.)  The ice surface in North Pembrokeshire might have been as high as 200m -- but it was probably no higher than that.

The directions of ice movement picked up from striations and other traces were from the NW (in western Pembrokeshire) and NNW in north Pembrokeshire.  This is consistent with the position of the ice edge.

However, we now have the problem of the so-called "Scilly Isles Surge" which carried ice all the way to the Isles of Scilly some time between 23,000 and 20,000 BP.  Moraines were laid down there just a little above present sea-level.  How can that have happened if the surging glacier responsible was the Irish Sea Glacier?  Over a distance of 200 km, the "surging glacier" must -- according to the hypothesis of James Scourse and others -- have had a long profile which was totally flat or a gradient of zero.  I don't like to use the word "impossible" -- but I hope that all who know even a little about glaciology might agree that whatever the ground sliding conditions and deformable bed characteristics might have been in the dry Celtic Sea at the time, this scenario is vanishingly improbable........

Things get even more improbable when we realise that there is a deep trench in the middle of St George's Channel which must have forced basal ice to flow UPHILL on its passage towards the south and south-west.  Glacier ice can only flow uphill when there is a sufficiently steep surface gradient, and a sufficient thickness of ice, to maintain forward momentum -- one again, if we are talking about the Irish Sea Glacier, that argues for thick ice in the channel, and not ice just 200-300m thick.

To return to the models in the literature that seek to explain the presence of Devensian ice in the Scilly Isles.  These three will illustrate what is in the literature:

The top diagram is from O'Cofaigh and Evans 2007, and the second and third come from Clark et al 2010.  The precise date for this "surge" is immaterial -- but it was probably sometime between 23,000 and 20,000 years BP.

So if these models are inherently improbable, what model will more reasonably explain the evidence on the ground, at least on the eastern flanks of the glacier involved?  Here is my suggestion:

The only scenario that seems to be consistent with the facts is one in which the Irish Sea Glacier builds up in an early phase of the glaciation but is then suddenly starved of its supply, and is replaced in St George's Channel and the Celtic Sea by a Celtic Sea Piedmont Glacier supplied by ice from the uplands of Southern Ireland.  There may have been a substantial ice dome there, with an ice surface altitude in excess of 600m.  This explains both the onshore ice directions in Pembrokeshire, and the presence of the glacier terminus at near present sea-level both in the Scillies and in West Wales.  What I am less sure about is the glacial sediment sequence in SE Ireland; and I also wonder whether a reconstructed piedmont glacier of this sort, when we work out its surface contours, will make sense with respect to glacier modelling and "ground truthing."

To the best of my knowledge, this is the first time that a piedmont glacier of this sort, in the Celtic Sea, has been proposed.  OK-- all you geomorphologists and glaciologists out there who are working on the LGM,  let's have your opinions!


Henry Patton said...

I agree that for Irish Sea ice to have extended down to the Scilly Isles seems theoretically improbable - Alun's last BIIS modelling paper (2009) showed that under this proposed scenario, great swathes of SW England would need to be inundated as well.

The idea of ice originating from Ireland instead does seem more likely, and has been mooted before (though nothing published as far as I'm aware). More extensive offshore mapping would be needed to prove such an idea, and it's something I hope to investigate with numerical modelling as well in the near future. Certainly climate parameters such as the precipitation distribution would need to be reconsidered.

BRIAN JOHN said...

Yes, it will be interesting to see where the modelling takes us on this.

In my post for 16 December 2010 I showed one of the "extreme" BRITICE models, showing the great (assumed) Irish Sea ice stream and ice covering most of SW England.

If we are to maintain respect for glaciological principles, the long thin surging lobe reaching the Scillies in spite of having virtually no long profile gradient looks very dodgy indeed...

BRIAN JOHN said...

Here is a message from Chris Hunt, which he will not mind me sharing:

I am not into blogging and my knowledge is old, but SE Ireland has historically been thought to have been beyond the LGM glacial limits.

However, I like the idea of dominant ice movement southeastward in the St Georges channel from a major ice-dome over Ireland - years ago I wrote a short paper on till palynology (Journal of Micropalaeontology, 1984) which included a section on material from the Fremington 'Till' (maybe proglacial lake clay, but certainly glacial) in N Devon. There was very good evidence, from the erratic palynomorphs, for the ice sheet moving in a southeastward direction onto the coast of SW England.

Similarly, White Limestone from Ulster, with a characteristic foraminiferal fauna, was found in the glacial deposits at Kenn, N Somerset, by Dave Gilbertson, again suggesting SE movement.

These two occurrences are from one or more ancient ice sheets on a much greater scale than the LGM one, but again point to the dominance of an Irish ice sheet, whenever that was...