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

Thursday, 21 April 2011

Did the last Irish Sea Glacier have a shallow long profile?

 With reference to this link:
A Glaciological Dilemma
I have been pondering on ice surface gradients.
When glaciers are in an equilibrium state, for example at the peak of a glacial stage with the ice at its maximum advanced position, they should approximate to the theoretical equilibrium profile described (and calculated) by glaciologists over many years.  The diagram above shows the theoretical profile (solid black line), dependent upon assumptions about basal shear stress, ice movement mechanisms, ice temperatures etc.  The profile approximates to a parabola.  Actual observed profiles on big glaciers, ice sheets and ice caps tend to be very close to the theoretical, showing that the theory is more or less correct.  We can see this on the diagram, with respect to Greenland and Antarctica.  But by definition, the glaciers we are observing today are either high-latitude glaciers, often cold-based and flowing from massive ice sheets; or mountain glaciers, heavily influenced by terrain.  Did glaciers during the main glacial episodes behave in the same way in the lowlands of the mid-latitudes, where there may have been very different thermal regimes in the ice, and where sliding on the glacier bed (for example) might have been very much greater?

Have a look at these long profiles of North American glaciers in mountainous terrain:
We can see that all of them have pretty steep long profiles, conforming quite closely to the theoretical equilibrium profile (ie a steepening towards the snout) but with quite strong regional topographic controls as well.  So, for example, Milne Glacier drops from 1000m to sea level in about 50 km; Henrietta Naismith Glacier drops 1200m in about the same distance; Bent Glacier is very steep, dropping more than 1000m in 30 km; and Gilman Glacier drops 1200m in about 45 km.  Click to enlarge the illustration.

At the time of the maximum (Anglian?)  glaciation of SW Britain, I'm still pretty sure that something as portrayed in the above map happened, with thick ice flowing S, SE and SW from St George's Chennel and with an ice surface around 2,000m between Fishguard and Rosslare.  That still gives a relatively shallow gradient, with a surface drop of 2,000m in c 500 km or 1,000m in about 200 km close to the ice edge.  (As we can see from the theoretical profile above, the ice surface gradient gradually declines with distance from the ice edge.)

But the situation in the Devensian Glaciation, around 20,000 years ago, appears to have been very different.  As noticed in my earlier posts (including the link above) a number of geomorphologists have recently suggested that there was a long shallow lobe of Irish Sea ice (with a gentle surface gradient) that pushed south-westwards from St George's Channel to the Scillies and beyond, at the peak of the Devensian glaciation:


I have been a bit sceptical about this, on the basis that ice with such a shallow gradient, with no topographical constraints, cannot have extended southwards as a long narrow lobe over 400 km long without spreading laterally onto the coasts of Somerset, Devon and Cornwall.

I still go for the Celtic Sea Piedmont Glacier idea, as portrayed on this map:

That having been said,  and having looked at some evidence from the Wisconsin glaciation of the Great Lakes area of North America I'm increasingly convinced that the gradient on this piedmont glacier might have been very shallow indeed.  North American geologists and geomorphologists have known for many years that near the southernmost extent of the Laurentide Ice Sheet during the Wisconsin, the ice flowed in a series of semi-independent lobes, reaching their maxima at different times.  Big end moraines and lateral moraines traced on hills and nunataks suggest that the lobes had incredibly shallow profiles, as shown on the top diagram with the green and red lines. (Excuse the fact that the lines are wobbly -- my shaky hand!!)   In Minnesota, South Dakota, Iowa and Montana the long profiles of these lobes was so shallow that the ice surface rose only about 500m over a distance of 500 km.  Now according to classic glaciological theory the ice should not have been capable of moving all that way with such a shallow surface gradient -- UNLESS the ice was moving rapidly in a vast surge, with a highly lubricated bed.  In other words, the shear stress in the glacier ice must have been vastly different from that of the cold-based glaciers used in the classic glaciology studies.  there must have been huge volumes of meltwater moving on the glacier bed......

So if the ice that crossed the Celtic Sea, moving broadly towards the SE from Ireland, was terminating close to present-day sea-level all the way from North Pembrokeshire to the Scilly Isles, what was the ice surface altitude on the coast of Southern Ireland?  I will hazard a guess and suggest that it was around 300m, but that's based on nothing but intuition and geomorphological instinct!

All contributions on this one gratefully received -- especially from geomorphologists.


Constantinos Ragazas said...


I am pleased, almost ecstatic, that you are now considering long surface gradients and glacier profiles! May I also suggest that the skeletal remains of an Alpine man found at Salisbury Plain may have been carried there embalmed in an ice-block by glaciers?

Soon, you will be discussing my 'ice transport theory' you have so adamantly rejected before!


Tony Hinchliffe said...

The Amesbury Archer was not carried to Salisbury Plain embalmed in an ice-block by glaciers. The Archer and his grave goods are all demonstrably from the date 2300 BC, when no glaciers were affecting Britain. See the Wessex Archaeology website for fuller analysis.