Some of the ideas discussed in this blog are published in my new book called "The Stonehenge Bluestones" -- due for publication on June 1st 2018. After that, it will be available by post and through good bookshops everywhere. Bad bookshops might not have it....
To order, click

Monday, 14 November 2016

The South Pembrokeshire problem

There's a problem in South Pembrokeshire.  It's probably not going to make the headlines, but it's quite an interesting problem if you happen to be a geomorphologist or an archaeologist, because it relates to the availability of land for settlement back in the Palaeolithic, around the time of the Devensian maximum ice advance. 

Over the years, I have published scores of maps on this blog, and most of them show an Irish Sea Glacier ice limit running along somewhere to the north of Mynydd Preseli and then looping down towards the mouth of Milford Haven (there are fresh glacial deposits at Druidston, West Dale, Mullock Bridge and West Angle) and then looping off again into Carmarthen Bay.  Here are a few examples:

You'll notice that in all of these maps, central and south Pembrokeshire are shown as being ice-free at the Devensian maximum, around 20,000 - 18,000 years ago.  The modelling of the waxing and waning of the Welsh Ice Cap, done by researchers in Aberystwyth University, has reinforced this assumption, showing Welsh ice pressing westwards in to eastern Pembrokeshire and with an intermittent Preseli Ice cap to the west interacting / coalescing with the ice of the Irish Sea Glacier when it was at its most advanced position.

But there are big problems with this scenario, which have been keeping me awake at night. (Well, not really, but it's a nice expression......) 

First, in the majority of published maps the ice edge is shown in a highly idealised form, sometimes as a virtually straight line.  Ice edges are not like this -- glaciers always butt up against obstacles and flow into lowlands, creating a fingered ice margin except in the rarest of circumstances.  So if the ice flowed in from the north-west and had a margin at an altitude of c 250m on the northern flank of Preseli, why would it have left the lowlands of the St David's Peninsula and central and southern Pembrokeshire ice free?  Clearly the ice would have flowed across these areas unless there had been an exceptionally steep ice edge -- and Celtic Sea researchers are now coming to the view that the ice surface gradient was remarkably low rather than remarkably steep.   Just look at the topography of Pembrokeshire and you will see that there is effectively no obstacle to Irish Sea Ice sweeping right across the county from the NW in the Devensian:

Second, the recording of unconsolidated till containing striated Old Red Sandstone cobbles and resting on carboniferous Limestone near the eastern tip of Caldey island means that Devensian ice must have crossed the island, travelling from west to east.  That means that the ice must also have crossed the Castlemartin Peninsula, even if actual traces of ice action on the limestone plateau are extremely rare.

Third, we now know that there are glacial deposits at around 340m on the northern flank of Preseli, not far from Tafarn y Bwlch.  It is reasonable to assume that the ice edge was at a similar altitude all the way eastwards towards Foel Drigarn and Frenni Fawr.  With ice movement from the north-west -- as demonstrated by the passage of the Broad Haven "super-erratics from Ramsey Island -- the ice must also have inundated the land surface of the western Cleddua drainage basin at least as far inland as Rosebush and Haverfordwest.  As it happens, fluvioglacial deposits are recorded from the Rosebush area, although I have not seen them myself.  When I put up a previous post on this, in 2014, I suggested that the deposits in the Rosebush sand pit had come from the Preseli Ice Cap -- now I am not so sure:

I have also interpreted the gravels at Llangolman as "probably Anglian in age" -- and I'm no longer so sure about that either! 

So should we now radically revise our opinions on the extent of Devensian ice in Pembrokeshire?   Well, maybe -- but we still have anomalies to cope with, including the remarkably fresh and fragile appearance of the tors at Maiden Castle, at the edge of the Trefgarn Gorge.  All of the fieldworkers in this area have assumed that they are so fresh that they cannot possibly have been overridden by Devensian ice:

 In contrast to Maiden Castle, the other tors and monadnocks of North and West  Pembrokeshire are heavily denuded and ice smoothed..........

Now another thing comes into the frame.  In a previpus post ion the Isles of Scilly, I referrd to an interesting article by Prof Danny McCarroll:

Trimline Trauma: The Wider Implications of a Paradigm Shift in Recognising and Interpreting Glacial Limits.  Danny McCarroll, Scottish Geographical Journal, 2016
Published 27 Feb 2016 

Trimlines mark the boundary between glacially eroded landscapes on low ground and landscapes dominated by evidence of periglacial weathering on higher summits. For many years the trimlines of Scandinavia, Britain and Ireland have been interpreted as marking the surface of the ice sheets at the maximum of the last glaciation, but recent cosmogenic exposure dating of erratics far above the trimlines in NW Scotland shows this to be false. The trimlines in that area must represent an englacial thermal boundary between warm (eroding) ice and cold (protecting) ice. It is now clear that even very experienced geomorphologists cannot necessarily tell the difference between terrain that has been recently glaciated and terrain that has not, because cold-based ice can leave virtually no trace. This calls into question not only the interpretation of high-level trimlines elsewhere, but also the mapping of the lateral limits of past glaciations, which are often based on similar or even weaker geomorphological and sedimentological evidence.

Danny refers to several of the areas which he knows well, and he says this about Pembrokeshire:

The ice limits that I know best are those in South Wales, and I now have very serious doubts about their veracity. In north Pembrokeshire, for example the proposed ice limit runs along the northern flanks of the Preseli Hills, leaving southern Pembrokeshire ice free. The evidence used to define that limit is exactly the same as the evidence we used to define the ice surface at trimlines. It is marked by a transition from bedrock that has been glacially scoured, and where glacigenic sediments are widespread, to a landscape dominated by blockfields and tors where glacigenic sediments are absent (Walker & McCarroll 2001). However, there are plenty of erratic boulders well to the south of the proposed ice limit, just as there are erratic boulders above the trimlines in Scotland. My best guess at the moment is that the north Pembrokeshire ice limit probably is the southern limit of the last ice sheet in that area and that the headwaters of the Cleddau remained ice free. That explains why that is the only river system in Wales that remains graded to well below the present sea level, producing the deep water port of Milford Haven. However, on a recent visit to the Castlemartin Peninsula, the limestone area south of Milford Haven, I noticed that the Carboniferous limestone is littered with erratic pebbles. There are also old records of large erratic boulders perched on the limestone, though most (perhaps all) have since been moved. Of course the erratics may have been deposited during an earlier glaciation, but I am not aware of any clear evidence to that effect. The concept of the ‘ice-free enclave’ of South Pembrokeshire should really be critically tested.

I am rather happy to go along with the sentiments expressed, although Danny is wrong to suggest that glacigenic sediments are absent outside the conventional Devensian limit.  I don't agree that the headwaters of the western Cleddau were ice-free in the Devensian -- I demonstrated relatively recent glaciation in the catchment in my doctorate thesis way back in 1965.......  And I am not sure what the depth of Milford Haven has to do with the Devensian -- it is clearly an immensely old inherited feature. 

But yes, let's really test the idea of this "ice-free enclave" -- and the evidence is stacking up that ALL of Pembrokeshire was probably affected by Devensian ice belonging to the Irish Sea Glacier, the Preseli Ice Cap, or the Welsh Ice Cap.  Watch this space.  In the meantime, here is a map that represents my current thinking on the position of the Preseli ice margin around 20,000 years ago:

 The thing that has really pushed me to this interptetation is the discovery of till not far from Tafarn y Bwlch when I was up there, wandering about with Chris a few months ago:

This till has come from the north, not from the south -- so it has nothing to do with a Preseli Ice Cap.  

Notice that I have postulated two ice lobes pushing through the mountain ridge via cols in the upland ridge.  The first of these, near Crymych, has already been suggested by the field surveyors of the BGS, and by other researchers.  The other one, in the Mynydd Bach col, is just to the west of Carn Goedog.  If there was an ice surface at around 340m, a lobe might well have pushed through.  But there is a rather interesting ridge-like feature right in the gap.  I've never noticed it before.   Is it a moraine?  As soon as I get a chance, I shall be up there to check it out..........


Dave Maynard said...

I was looking at The Incredible Human Journey - 5. The Americas, on Iplayer.

Alice Roberts was discussing the Devensian glaciation of Canada and Alaska and showing a graphic of its extent. Obviously this was very generalised, but it showed the northern edge of the Alaskan archipeligo as ice free. Is this so? I'd have thought it would have been completely covered and the Bering Strait one block of ice. There may be some reason for this as opposed to a dumb graphic artist.


BRIAN JOHN said...

The high Arctic is not the best sort of environment for big glaciers / ice caps / ice sheets to develop. Not dynamic enough. In the big glacial episodes, parts of the high Arctic are so cold and dry that there is insufficient snowfall to maintain glaciers -- so there are "Arctic deserts" and tundra instead. Further south, around 55 deg to 60 deg N, where winds are coming off an ice-free ocean, vast amounts of precipitation can cause ice sheets to grow quite rapidly, especially when positive feedback mechanisms come into play. But you can get arid and ice-free areas on the lee side of these big ice masses too.

Most people think that most of Alaska was ice free during the Late Wisconsin glaciation. What is of greater interest to us is the "driftless" area of the American midwest, which may or may not have completely escaped glaciaton!