Following the contribution from Barry, let's have a look at the central hypothesis here. The maps above come from Etienne et al, and are based on detailed work in the lower Teifi Valley and more scattered bits of work elsewhere. Their main contribution to glacial geomorphology was to show that the the lake deposits (rhythmites or varves of clay and silt) up-valley from Cardigan were laid down in glacial lakes at a time of ice advance. That's a very different scenario from that of Charlesworth, who argued in 1929 that there were great lakes in this area AFTER the peak of the last glaciation, at a time of ice wastage. That's the conventional or accepted scenario, and it's what happens most often in ice marginal situations as well.
So what the new work means is that if there are tills and fluvioglacial deposits in the area, dating from the Devensian glacial episode, they should be ON TOP OF the lacustrine deposits. That indeed seems to be the situation in the Teifi Valley. But what about the rest of the area shown on the maps above? There have been hardly any systematic borehole investigations, and there are very few natural exposures we can look at -- so what we have in map (f) above -- the one enlarged in the lower illustration -- has to be looked on as a working hypothesis, to be tested against field evidence.
For a start, I have already mentioned on this blog that I'm not that happy with either the ice directions portrayed on the map, or with the straight smooth nature of the ice edge. Ice edges in this sort of undulating terrain look much more like this:
So unless we can find evidence of lake deposits in the areas shown as being occupied by water in the lower map, we should keep an open mind. Maybe the area of Lake Brynberian was actually occupied by a lobe of ice rather than by a lake?
There are two critical altitudes here, if we follow the hypothesis of the researchers. One is the 220m water level, shown on the map, with a spillway at 220m via the Rhosddu Channel near Crymych. The other -- at an earlier stage -- is at 115m, with a spillway into the Gwaun Channel at Cilgwyn. (Other lower spillways are also postulated for stages during which the ice front was out in Cardigan Bay, ponding meltwater against the coastline.)
So we are talking about two hypothetical lake shorelines, one along the north face of Preseli, below Carn Goedog and Carn Alw; and another much lower down, in the currently farmed area, coincidentally running quite close to Rhosyfelin, running through Tycanol Wood, and then running through to Cilgwyn. Unless these lake levels were VERY short-lived, we might expect to see some evidence of them in the landscape. We need to look more carefully. But at the moment I have reservations, since I do not see a clear spillway at Cilgwyn (which is where I live) and instead I see a large moraine, with banks of sand and gravel which seem to me to be kame terraces rather than lake deposits.
With respect to the sequence at Rhosyfelin, I think that the digging team in 2012 has got down to the top of a layer which seems to me to be till. They refer to it as a "surface" or floor, and clearly think that this was the floor of the quarry on which all those heroic quarrymen worked back in the Neolithic. I think it is a sedimentological rather than anthropological feature -- although I would have no problem with the lower part of it, near the tip of the rocky outcrop, being used in Iron Age times (or whatever) as a camp site with a hearth.
So how do we fit the Rhosyfelin sequence together? There may be lake deposits under the till. Only time will tell. The fine-grained material which seems to be above the till might represent a short-lived lacustrine situation, or it might be a "redistribution" layer, made of older lake deposits from upslope which have carried down towards the valley floor. Then we have that thick sequence of slope deposits, partly periglacial and partly seeming to indicate warmer climatic conditions.
Where does all the shattered rock debris come in? I think it might rest on the till -- that would make most sense to me, given that we have had 10,000 years of oscillating cold climate following ice retreat and culminating in the "cold snap" of the Younger Dryas.
We eagerly await the results of the pollen analysis and radiocarbon dates from the excavated sides of the dig, to tell us just how large the time-span is between the current ground surface and the "floor" of the 2012 excavation.