Those Meltwater Channels

The Gwaun - Jordanston System of meltwater channels in North Pembrokeshire -- possibly the most impressive suite of subglacial channels to be found anywhere in the world.  Several of these channels are cited in SSSI citations, and since they are deemed to be "classic landforms" they are carefully protected within the Pembs Coast National Park.    The Gwaun Valley is the most spectacular -- that's the one which runs in a long loop from the right-hand (east) edge of the images and out at the coast at Fishguard -- effectively cutting off the Carningli-Dinas Mountain massif from the rest of the Preseli uplands.

The image above (courtesy Jonathan Lee) shows the channels wonderfully well -- click to enlarge and see more detail.

The orientation of these channels has always been problematical, because they suggest that at the time of formation ice was flowing across the landscape from NE towards SW.  That's not at all the direction of ice movement which we pick up from the erratic trains and striae in Pembrokeshire fieldwork.  But there are some striations in the St David's area (see previous post) that suggest ice movement from NE towards SW, even out at the tip of the peninsula.  That means unhindered Welsh ice flowing out into St George's Channel at some stage.  When might that have been?  Well, I think the channels are very old, since they contain lodgement till and fluvioglacial deposist probably dating from the Devensian.    My interpretation of the age of these channels is very different from that of Prof David Bowen and others, who have thought of them as Devensian features associated with ice wastage maybe after 23,000 years ago.  But there are major problems with that dating.  So my interpretation is that the channels were formed maybe in the Anglian Glaciation, at a time when Welsh Ice was dominant -- and when there was no Irish Sea Glacier occupying St George's Channel.  It's most reasonable to assume that this was in an "early Anglian"phase, after which the Irish Sea Glacier expanded and established an ice flow across the channel complex, with ice moving from NW towards SE.  So the channels would have been plugged with stagnant ice while more active ice sheared over the top of them.  Parts of the channels would have been used and refreshed (and altered) during deglaciation, and then again during later glacial phases including the Devensian.

That's my story for now, until somebody comes along and convinces me that the evidence supports another one!