I have come to the conclusion that we have all been getting it wrong. The edge of the Irish Sea Glacier in west Wales during the Devensian Glaciation or LGM was not wrapped around the Pembrokeshire coastline at all, but was much further east. In fact, I now think that the whole of Pembrokeshire was inundated beneath glacier ice around 26,000 years ago. This was all Irish Sea ice derived from far to the north, and from Ireland, but I'm still attracted to the idea that there was a stagnant or sluggish Preseli ice cap over the highest parts of the upland.
I have suggested this before, albeit with not much conviction.......
You can find many other posts by using the excellent search facility on this blog.
The ice edge maps in the literature are now so out of date, and indeed nonsensical from a glaciological standpoint, that they are best consigned to the dustbin. Here are some of the dodgy ones (including some of mine!):
Recent work on the glaciation of the Celtic Sea shelf by the Irish Sea Ice Stream has shown that the Devensian ice surface gradient was exceptionally shallow, dropping from +750m in the Irish Channel to +400m in St Georges Channel to -200m at the southernmost ice edge about 500 km away. That represents a surface drop of about 1000m over a distance of 500 km -- or 2m per km. There must have been substantial lateral spreading and gradient reduction on the ice surface once it was clear of the constriction between North Pembrokeshire and the coast of SE Ireland. It was proposed by Prof Geoff Boulton many years ago that this was because of the soft and saturated sediment bed on what had been the sea floor, facilitating a high rate of bed deformation, lubrication and basal sliding. This has been confirmed by subsequent work by James Scourse and the BRITICE team. So could the ice from the west have surmounted the chalk escarpment on the edge of Salisbury Plain? Since the Plain has a surface altitude of 200m-250m, an ice front capable of surmounting the chalk scarp must have had a surface altitude of c 300m. That could not have happened in the late Devensian glaciation, but it would have been perfectly feasible in the Anglian, at a time when the whole of Preseli must have been deeply inundated by ice, with a glacier surface at c 750m. The soft sediments of the Somerset Levels would have facilitated ice movement eastwards from the Bristol Channel. The drop in the surface of the ice lobe would have been 450m over a distance of 280 km -- namely 1.6m per km. That's a reasonable ballpark figure, given that the deforming sediments in the Bristol Channel must have been very similar to those of the Celtic Sea, in both the Anglian and the Devensian glaciations, leading to a very similar style of glacier behaviour. I will look at this topic again, in another post.