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

Friday 31 August 2018

More South Pembrokeshire till locations

Look almost anywhere, and thou shalt find.......

I have been looking at the Tenby and Pembroke geological memoir, edited by E.E.L.Dixon and published in 1921. This was of course the year in which HH Thomas gave his famous lecture on the Stonehenge bluestones, and denied the possibility that the monoliths could have been transported by ice.  Thomas was one of the geological surveyors, and some of the notes in the chapter on superficial deposits may well have been written by him.

That's interesting, because in this memoir, as in the others relating to West Wales, a vast amount of evidence (collected by Thomas, Cantrill, Strahan, Dixon and Jones) is presented which shows that glacial deposits are widespread, and that erratic transport was not just possible but well documented.  In the memoir, there is reference to erratics transported from Pembrokeshire to the Swansea area and to Pencoed in Glamorgan.  Thomas knew that boulders could be carried by ice for at least two hundred miles -- and indeed he was perfectly familiar with big erratics of Scottish origin in Pembrokeshire -- and yet he deemed the transport of Stonehenge bluestones to have been impossible.  So his bluestone lecture, and the paper that followed it in 1923, was designed for effect -- presented to antiquarians, not geologists.  He wanted to tell a story, and I think he wanted notoriety and fame.  Now where have we heard that before?

Ixer and Bevins have done exactly the same thing, of course, publishing the bulk of their recent articles about the bluestones not in the specialist geological literature, but in magazines read by archaeologists and the general public -- edited by archaeologists and presumably refereed by archaeologists too.  Limited scrutiny, instant appeal......... and uncertain reliability.

On to the till.  The surveyors do not always identify glacial deposits for what they are in in these memoirs --  we must recall that the term then in use was "boulder clay", not "till", and when described deposits were not particularly clay-rich, they often referred to them as "gravels" with faceted and striated erratics in them!  I am quite certain that many of the deposits called "gravels" are in fact of glacial origin.

Caldey Island -- Sandtop Bay.  On page 196 there is a record of what has to be fresh till, in a deep pipe in the NE corner of the bay.

Lydstep Haven.There is a record (on page 201)  of the submerged forest which was exposed in 1906, with a basal stony deposit under it, containing limestone and sandstone pebbles, chert and igneous rocks.  The surveyors speculate that this material is "head", "mixed with some Drift debris".    That sound to me like a till which has  been subjected to some redistribution under cold climate conditions prior to the growth of the submerged forest.   This is the classic Late Devensian scenario.

Low-res photo from David Evans, Pembvs Coast Photography, of the Black Mixen peninsula.  The blackish lichens or algae on the rock surface probably explain the name.....

Draught, at the exit point of the Lydstep Valley, is referred to on p 200 as having a small patch of gravel and sand -- "evidently ancient" -- containing chert, grits, shales and ORS pebbles.  The surveyors say it was being eroded. at the time of the survey.  They say it is probably nothing to do with the raised beach because it has no head above it -- and I am convinced this is the first record of the Devensian till which I described in a recent post.

On p 192 there is mention of the "Black Mixen" peninsula at Lydstep, which has to be the one which has the ancient till exposed on its neck, beneath cemented limestone breccia.   Strangely, the surveyors make no mention of the cemented till.

Freshwater East -- there is a record of "strips of gravel" a few feet above the alluvium of the river valley.  Origin uncertain.

On pp 199-200 there are many mentions of erratic boulders, including the seven at Flimston Church, which have been commandeered for use as headstones.

Stackpole Quay -- on p 198 there is reference to a patch of "gravel" incorporating many erratics, including igneous ones, at c 100 ft altitude at the top of the cliff of the "north cove".  It occupies a notch, and seems to have a cap of head above it.  According to the record, the gravel (it must be till) is 10 ft thick, and the gravel above it is 8 ft thick.

Bosherston Limestone plain -- drift with chert, ORS and igneous pebbles here and there von the surface and sometimes in limestone "pipes". Some igneous erratics are over 18" in diameter.  

Flimston clay pits -- the surveyors remark upon the Tertiary quartz gravels with many white pebbles, but also upon the presence of "red gravel" containing ORS pebbles, chert, and faceted and striated pebbles including many of igneous origin.  Again, this must be a relatively fresh till.

Freshwater West -- submerged forest north of Little Furzenip exposed in 1912, with peat and prostrate branches, stumps and roots, the latter embedded in "head".  There is a stony red clay which is bleached to green-grey near the surface. Stones in this clay include igneous erratics.  This deposit is separated from the submerged forest beds by a tenaceous blue clay up to 6" thick.  (interestingly enough, the forest here has traces of human occupation, and it also contained the remains of a whale skull which is apparently preserved at the National Museum of Wales (Whales?)........

See also:

The Geology of the South Wales Coalfield. Part XIII: The Country around Pembroke and Tenby. Being an account of the region comprised in Sheets 244 and 245 of the map. By E. E. L. Dixon, B.SC, F.G.S. Memoirs of the Geological Survey, pp. 220. 1921.

Thursday 30 August 2018

Rising and sinking coasts; isostatic and eustatic effects

Inishowen, County Donegal, Ireland.  A rising coastline.......

Another big and important paper which will help us to elucidate what went on during the Late Devensian.

Relative sea-level changes and crustal movements in Britain and Ireland since the Last Glacial Maximum
by Ian Shennan, Sarah L. Bradley, Robin Edwards
Quaternary Science Reviews 188 (2018) 143-159


The new sea-level database for Britain and Ireland contains >2100 data points from 86 regions and records relative sea-level (RSL) changes over the last 20 ka and across elevations ranging from ~+40 to -55 m. It reveals radically different patterns of RSL as we move from regions near the centre of the Celtic ice sheet at the last glacial maximum to regions near and beyond the ice limits. Validated sea-level index points and limiting data show good agreement with the broad patterns of RSL change predicted by current glacial isostatic adjustment (GIA) models. The index points show no consistent pattern of syn- chronous coastal advance and retreat across different regions, -100-500 km scale, indicating that within-estuary processes, rather than decimetre- and centennial-scale oscillations in sea level, produce major controls on the temporal pattern of horizontal shifts in coastal sedimentary environments.   Comparisons between the database and GIA model predictions for multiple regions provide potentially powerful constraints on various characteristics of global GIA models, including the magnitude of MWP1A, the final deglaciation of the Laurentide ice sheet and the continued melting of Antarctica after 7 ka BP.

This is a very impressive data collection and analytical exercise, and this is the key map:

Click to enlarge. What this map shows is five different RSL regions, all affected by the same eustatic sea-level rise since the end of the last glaciation, but all with different histories of isostatic and tectonic adjustment. The brown dots (sampling areas) mark the area of greatest isostatic uplift associated with the melting of the Celtic Ice Sheet;  this was the "core area" and the last to melt.  Overall, in this area, RSL has fallen -- see the diagram at top left.  The area with the yellow dots was also heavily inundated by ice and shows a similarly complex RSL history, but with isostatic recovery rates more or less in step with the eustatic sea level rise.  The area with the green dots experienced some isostatic uplift (the ice was thinner and not so persistent)  but an overall rise in RSL until the levelling off which started around 7.000 BP.  In the area with blue dots, although there was intense glaciation in places, the ice cover was relatively thin and short-lived, so that the amount of isostatic depression was quite small,  leading to a complex interaction between eustatic and isostatic effects up to about 10,000 BP and a relatively straightforward RSL rise since then.  The area with the black dots was mostly outside the range of the Devensian glaciation, so although there were some isostatic "forebulge" and other effects, the RSL rise has been relatively straightforward and gradual since 20,000 years ago.  This can all be seen in the graphs attached to the map.

This is another usefil diagram, showing the predicted eustatic sea-level curve for the last 20,000 years, commencing at about -120 m and rising irregularly since then, with an additional kink about 7,000 years ago for continuing Antarctic ice melt.

The top curve on the graph is a proxy for the climate change curve.  The bottom line with the black dots shows the predicted eustatic (global) sea level rise, and the other thin lines show the RSL curves for the 86 regions studies and marked on the map.

And one more very useful diagram:

This one separates out the isostatic and eustatic components.  It shows the same sea-level curve, but shows the range of isostatic recoveries demonstrated for the 86 regions -- as we can see, the amount of isostatic adjustment varies between zero and about 150m.  For West Wales, the amount of recovery is probably less than 50m, and this is suggestive of relatively thin ice.  I must get the actual figure from the authors of this paper..........

Overall, very useful indeed.  The paper will be much cited.

Holocene raised beach ridges near Tarbert, Isle of Jura.  In this area the highest strandline 
is around 40m.

Exposed Holocene raised beach ridges near the NW coast of Jura -- seen on a satellite image

Another satellite image of a "washboard" of raised beach ridges, on the west coast of Jura

Wednesday 29 August 2018

Is there a Devensian terminal moraine south of Caldey Island?

As all faithful blog readers know (I was tempted to write, in homage to my guru: "... as any fule kno....") I have been banging on about the Devensian glaciation of Caldey Island for ages.  Just stick "Caldey Island" into the search box and see what comes up.  But on looking through some old files I came across this map, showing the nature of the sediments on the floor of Carmarthen Bay.  Very interesting indeed!

As we can see, to the south of the island there is an extraordinary mixture of sediment types, ranging from mud to sandy mud to muddy sand to assorted types of sand and gravel and even "muddy sandy gravel" -- which sounds to me suspiciously like till..........  Is this all that is left of a long looped ridge that might have been a terminal moraine, marking the edge of the Late Devensian Irish Sea Ice Stream? There does not seem to be any trace of a physical feature on the sea bed today, but I like the thought......

The above map shows the reconstruction for the Bristol Channel landscape after the end of the Devensian Glaciation, during the late Upper Palaeolithic -- while the sea was starting to flood back in from the west.

If the ice edge was located somewhere around here, what of the other evidence?  Well, we have the apparent dilemma of till at the eastern end of the island, at Ballum's Bay, very close to one of the caves which has been investigated.  This is Nanna's Cave and Ogof-yr-Ychen, which I examined with Brother James and Mel Davies back in the good old days.

There are animal and human remains in the Caldey caves, mostly dating from between 20,000 and 30,000 years BP.  There are doubts about the accuracy of some of these old radiocarbon dates, relating to new protocols of correction or adjustment, but broadly these dates coincide with the peak of the Late Devensian glaciation.  So were all these wild animals prowling around the ice edge?  Were the caves overwhelmed by glacier ice at the peak of the glaciation?  Were human hunters in residence in these caves at the time that the ice advanced to its maximum extent and then started to retreat catastrophically?

We don't really know the answers to these questions, but maybe they are things we should now concentrate on..........

Paviland and its "RED LADY" (who was of course a man) also come into the frame, since Paviland Cave is not very far away, on the Gower Peninsula.

We need to piece together the geomorphological evidence with the evidence of Paleaolithic prehistory.  More thought needed......


PS.  3rd Sept 2018. 

I am more and more convinced that the Paviland Moraine in Gower is of Devensian age.  I have considered the evidence here:

Prof Dai Bowen argued that the moraine was of Anglian age.  Indeed, this dating was crucial in his denial that Irish Sea ice could have carried bluestone boulders to Stonehenge.  His reasoning was that if the Welsh Ice flowing southwards during the Anglian Glaciation skidded to a halt near the coast of Gower, Irish Sea ice could not have flowed across the Gower Peninsula towards Somerset and Wiltshire, and there can have been no mechanism for transporting the bluestones.

However, Prof DQB made a lot of dating mistakes in the course of his career, and I am convinced that this was one of them.  If the Paviland Moraine is Devensian it would be similar in age to the glacial deposits on Caldey -- and Anglian Irish Sea Ice could well have passed across this area on its way to Stonehenge.

Things are knitting together in a variety of ways, some of which are unexpected.......

Monday 27 August 2018

British Isles relative sea-level map

This is slightly off-topic, but worth sharing.  A beautiful map from Prof Ian Shennan and colleagues (Ian was in the Durham Univ Geography Dept when I was there, and he is still there).

The map shows current relative sea-level changes in the British Isles -- the result of a vast data collection exercise.  Note that in the north the land is rising slowly relative to the sea, and in the south it is sinking.  There are three basic factors:  glacio-isostatic adjustments still going on following the melting of the last ice sheet; crustal deformation as a result of tectonic forces; and the eustatic sea-level rise connected to global warming.

Main message:  if you live in the south of Ireland or the UK, you have a vested interest in stopping the rise of global temperatures.  Secondary message: if you are thinking about investing in some coastal real estate in the Isles of Scilly, think again.

PS. Sorry  -- should have given more details:

The units on the map are mm per year.  Minus symbol used for "sinking" areas.

Also:  Shennan, I., Bradley, S.L. & Edwards, R. Relative sea‐level changes and crustal movements in Britain and Ireland since the Last Glacial Maximum. Quaternary Science Reviews. 2018;188:143-159.

The three-dimensional jigsaw puzzle

They are coming thick and fast.  Another very interesting paper from James Scourse and colleagues, discussing the role of tides and relative sea-level in determining how the last Devensian ice sheet in the British Isles collapsed.  The article is not easy to read -- its language is very convoluted -- but it suggests that there was a major difference in the style of deglaciation / ice sheet collapse in the western fringes of the British Isles as distinct from the Norwegian Sea / North sea sector because in the former there were "mega-tides" and in the latter there were much smaller tidal amplitudes.  This is shown in this map, modelled for 16,000 years ago.

On the map we see tidal amplitudes of over 2m on the western fringes, but less than 1m in the north and east.

As we know, tidal range increases where there are topographic "funnel effects", as in the Bristol Channel of today, where in the inner reaches the tides have an amplitude of over 4m.  The extreme tidal range is in the Severn Estuary, where it is around 50 ft or 15m.  But at the peak of the last glaciation sea-level was around 120m lower than it is today, and so the coastal "palaeo-topography" was quite different. The coastline was out near the shelf edge, so while there may have been funnelling effects, they would have been in different places.   But that's not the whole story, since glacio-isostasy has to come into the equation;  as the ice sheet waxed and waned, so did its isostatic loading on the crust.  That would have changed the alignment of the coastline, with the sea flooding into depressed areas -- except where grounded ice prevented it from doing so.

A hugely complex three-dimensional jigsaw puzzle.  So good for James Scourse and other authors (in this paper and many others) for trying to sort it all out........


"The role of megatides and relative sea level in controlling the deglaciation of the British–Irish and Fennoscandian ice sheets

J. D. Scourse, S. L. Ward, A. Wainwright, S. L. Bradley, and K. Uehara
Journal of Quaternary Science: Volume 33, Issue 2
February 2018, Pages 139-149

Key external forcing factors have been proposed to explain the collapse of ice sheets, including atmospheric and ocean temperatures, subglacial topography, relative sea level and tidal amplitudes. For past ice sheets it has not hitherto been possible to separate relative sea level and tidal amplitudes from the other controls to analyse their influence on deglaciation style and rate. Here we isolate the relative sea level and tidal amplitude controls on key ice stream sectors of the last British–Irish and Fennoscandian ice sheets using published glacial isostatic adjustment models, combined with a new and previously published palaeotidal models for the NE Atlantic since the Last Glacial Maximum (22 ka BP). Relative sea level and tidal amplitude data are combined into a sea surface elevation index for each ice stream sector demonstrating that these controls were potentially important drivers of deglaciation in the western British Irish Ice Sheet ice stream sectors. In contrast, the Norwegian Channel Ice Stream was characterized by falling relative sea level and small tidal amplitudes during most of the deglaciation. As these simulations provide a basis for observational field testing we propose a means of identifying the significance of sea level and tidal amplitudes in ice sheet collapse.

Increases in tidal amplitudes……… promote increased ice stream flow rates, promote iceberg calving and destabilize ice shelves, and the de‐buttressing that results from ice shelf collapse promotes ice stream acceleration. Tides are therefore integral to a series of positive feedbacks that promote ice sheet collapse with implications for global sea level.

Several independent global palaeotidal model simulations are now available (Egbert et al., 2004; Uehara et al., 2006; Arbic et al., 2008; Griffiths and Peltier, 2009; Green, 2010) that indicate that the glacial North Atlantic, with ice‐equivalent sea levels 120–130 m lower than present (Peltier and Fairbanks, 2006), was characterized by megatidal amplitudes (amplitude > 5 m, tidal range > 10 m) and increased tidal dissipation rates (Griffiths and Peltier, 2009; Wilmes and Green, 2014).

Sunday 26 August 2018

Sand ridges on the Celtic Sea shelf edge

Celtic Sea linear tidal sand ridges, the Irish Sea Ice Stream and the Fleuve Manche: Palaeotidal modelling of a transitional passive margin depositional system
by James Scourse, Katsuto Uehara, and Adam Wainwright
Marine Geology 259 (2009) 102–111


The linear tidal sand ridges (LTSR) of the Celtic Sea constitute the largest examples of their bedform type on Earth. Previous sedimentological and seismic stratigraphic interpretation suggests that the LTSR are moribund tidally remobilised sediments representing the transgressive systems tract. This interpretation is supported by two-dimensional finite-difference model reconstructions of the M2 tide, forced using the output from a glacial isostatic adjustment model to derive palaeotopography, and a global ocean model to derive the tides on the ocean boundary, used here to reconstruct peak bed stress vectors for the Celtic Sea for timesteps covering the transgression since the Last Glacial Maximum (last 21 ka). These data are coupled with interpretation of recent published observations on the outer shelf depocentres of the Fleuve Manche and the Irish Sea Ice Stream (ISIS) to confirm that the LTSR distribution is consistent with modelled sand transport paths during transgression. The main phase of LTSR growth was between 20 cal ka and 12 cal ka. Ridge axis orientations reflect the final phase of LTSR construction around 12 cal ka, with some later growth of the most southerly LTSR as late as 10 cal ka. LTSR growth was from the SW across the shelf towards the NE. Strong tidal pumping of sediments into slope canyon heads on the outer shelf occurred between 20 cal ka and 12 cal ka, contributing to turbidite activity and the growth of the Celtic and Armorican deep sea fans. It is proposed that 1. the Fleuve Manche shelfal estuary-delta, and the Irish Sea Ice Stream (ISIS) shelf fan, were the main sediment depocentres supplying the growing LTSR, 2. the lack of sediments available in the western Channel limited LTSR growth in this area, and that the easterly termination of the LTSR in this sector results from sediment starvation, and 3. the northeasterly termination in the Celtic Sea sector is a function of declining peak bed stresses rather than sediment starvation. The Celtic Sea margin represents a passive margin depositional system transitional between true glacial ice stream-trough mouth fan systems (to the north) and the fluvial canyon systems characteristic of the margin to the south. This interpretation therefore complements studies of macroscale sedimentation linked to glaciation on the continental margins of the North Atlantic.

That is not the easiest Abstract to understand (nor is the paper), but the gist is that these big ridges in the Celtic Sea, out towards the shelf edge and in water depths of between 130m and 200m, are probably not glacially streamlined features or De Geer moraines, but the products of tidal “pumping” in areas with huge sediment supplies. The sediments shifted about might well have had glacial origins, since the Devensian ice certainly did reach the area to the west of the Isles of Scilly (and according to the latest paper by Praeg et al it probably reached the shelf edge. The idea is that the sediments were pumpted over the shelf edge and into a set of distributory channels which carried them down to the Celtic Deep Sea Fan and the Armorican Deep Sea Fan in deeper water. But I am still mystified as to why there are ridge forms — as the authors explain, some of the ridges are "as much as 200 km in length, 55 m high and 15 km wide, with a ridge-to-ridge wavelength of some 20 km”. I am not sure that the artice does this, but maybe somebody will explain to me why these features have elongated ridge forms and why we do not simpley see an unduating thick spread of sea floor sediments……….

Anyway, a very useful paper, with some interesting maps added. Figure 3 is fascinating, illustrating the peak bed stress vectors for the Celtic Sea in “timesteps” since the peak of the Devensian Glaciation. The evolving shoreline is shown; so we see that 21,000 years ago the Celtic Sea arena , Bristol Channel and St George’s Channel were dry land. That means that any glacier ice still present would have been grounded. By 19,000 BP some lakes were forming to the west of embrokeshire, and these get bigger and bigger until around 15,000 BP. Then the rising sea breaks through, and the centre of St George’s Channel has sea water flowing through it around 14,000 years ago. The shoreline creeps closer and closer to the westernmost cliffs of Pembrokeshire and by about 10,000 BP it is not far from the Pembrokeshire islands of Ramsey and Skomer. By 8,000 BP the water is flooding into St Bride’s Bay and Carmarthen Bay.

The other interesting map is Figure 2, showing not only the famous elongated ridges but also a reconstruction of the Devensian British and Irish Ice Sheet (or Celtic Ice Sheet, as it is now being called). I know this is a bugbear of mine, but I am intriged by the reconstructed flowlines. All of them look perfectly sensible, perpendicular to proposed ice edges — except for the Celtic Sea lobe, where everything looks cockeyed. The eastern edge of the lobe should be much further east — but we’ll forgive that, since this paper is almost 10 years old. But the flowlines should be fanning out, not converging, and ice should be flowing towards the Scilly Islands and the tip of Cornwall from the NW. Much as I admire geomorphologists, there seems to be a corporate mental blockage on this issue.

PS.  For previous mentions of this topic, use the search box.

More wisdom from Peanuts

With respect to the things we have been talking about recently, I came across this on a Facebook page and thought it might be a nice representation of an archaeology class at one of our best-known universities........

This is the final sentence of my "working paper" of August 2016:

"It is for the academic archaeological community itself to determine whether the telling of stories has now become more important than the objective collection of field evidence and the application of scientific rigour in the process of analysis and the drawing of conclusions."


Saturday 25 August 2018

The Carn Goedog Files

These are the four assessments which I have made of the recent paper by Parker  Pearson et al which is published in "Antiquity" journal.

The paper itself:

"Megalith quarries for Stonehenge’s bluestones", by Mike Parker Pearson, Josh Pollard, Colin Richards, Kate Welham, Chris Casswell, Duncan Schlee, Dave Shaw, Ellen Simmons, Adam Stanford, Richard Bevins & Rob Ixer. Antiquity 2018, issued as a pre-publication PDF by Southampton University (not in final format)

I have made many other comments on the Carn Goedog research over the past few years -- use the search facility on this blog to find them.

This paper is so seriously defective, in almost every respect, that I find it bizarre that it ever found its way into an academic journal published by Cambridge University Press.  It is not a research paper; it is piece of unabashed marketing.   I have asked this before, and I ask it again -- where is the scrutiny from within the archaeological establishment?  How is it that so many serious and senior archaeologists -- and two senior geologists -- have allowed their names to be attached to it as co-authors?

And the most serious issue of all.  If I, as a local person with a detailed knowledge of this site and with an academic background, had not been around at this moment in history, and had not been able or willing to look at the excavation site and to scrutinise the research output from the MPP team, everything in this article would have been accepted as THE TRUTH.  Just think about it..........

And think about this.  If this is the level of non-scrutiny applied to Rhosyfelin and Carn Goedog, how many other British archaeological sites are there which have been wildly misinterpreted and which have had  nonsensical narratives attached to them?

Rhetorical questions, I know.  But sometimes they are needed.  Actually, rhetorical questions don't need answers.  Maybe these questions do. 

There are plenty of sensible archaeologists around.  But when are they going to speak up?  Quite seriously, if they do not, archaeology will become a standing joke.

Friday 24 August 2018

More from the Megalithic Quarrymen (4)

This is the paper:

"Megalith quarries for Stonehenge’s bluestones", by Mike Parker Pearson, Josh Pollard, Colin Richards, Kate Welham, Chris Casswell, Duncan Schlee, Dave Shaw, Ellen Simmons, Adam Stanford, Richard Bevins & Rob Ixer. Antiquity, June 2018 .

The last two sections of the paper are entitled "Megalith-quarrying at Carn Goedog and Craig Rhos-y-felin" and "From the bluestone quarries to Stonehenge" .  These are the sections that would be called -- in a conventional academic paper --'"Interpretation" and "Discussion".    But this is no ordinary paper, and as as we have seen, the authors see no merit at all in presenting their evidence and considering any alternative explanations for the features described.  So with a degree of certainty which we continue to find  quite amazing,  they simply continue with their narrative.........

And it gets ever more convoluted and confusing, with more radiocarbon dates thrown into the mix.  Rhosyfelin is also brought into the narrative, with some dates previously unreported.

The authors start by saying:   "At least five bluestone pillars (Stones 33, 37, 49, 65 & 67) were taken from Carn Goedog, and probably many more (Bevins et al. 2013). The multiple and large recesses in the rock face are further evidence that pillar removal was extensive at this outcrop....."   I don't recall Bevins et al saying that  at least five bluestone pillars (with numbered identities) came from here;  their paper was much more circumspect,  and as I have pointed out many times, they did not demonstrate that any spotted dolerite bluestones actually came from the tor rather than from other places along the sill that extends towards Carn Alw.  But one doesn't want the truth to get in the way of a good story, does one?

As for Rhosyfelin, there is a change of tone.  In this article the authors claim that "at least one pillar" came from here, and maybe one or two more.  That claim is still, of course, completely speculative, and is unsupported by any evidence, as my two colleagues and I demonstrated in 2015.

The authors claim that "artificial platforms" were constructed at both of their putative quarrying sites, and that at Rhosyfelin there was a vertical drystone retaining wall at the platform edge. We have looked at this before, and have shown it to be entirely fanciful.  Equally fanciful is the "hollow way" cut into soft riverine sediment.  New radiocarbon dates of 4434 yrs BP, 4404 yrs BP and 4627 yrs BP have been obtained for charcoal fragments in the alluvium resting in the "hollow way".  Parker Pearson et al argue, on this basis, that the hollow way had gone out of use by this time;  but if you consider (as I do) that the hollow way is simply the product of a fertile imagination, the dates simply indicate that during the aggradation or accumulation of sediments on the valley floor there was possibly burning and forest clearance somewhere upstream.   That would not have been surprising, in an area full of Neolithic and Bronze Age settlement traces.  A so-called rhyolite end-scraper was found in the sediment fill on the platform, with two radiocarbon dates of around 6,000 yrs BP;  these are extremely inconvenient, and are therefore classified as being "residual in redeposited material".  Ah, if in doubt about something, call it a ritual feature, and if a date is wrong, call it residual........

In seeking to demonstrate a common history for Rhosyfelin and Carn Goedog, the authors do not make a very good fist of it.  They confuse the issue by using phrases like "the second half of the fourth millennium BC", but essentially they are saying that most of the prehistoric dates for Carn Goedog fall within the period 3350 - 3000 cal BC or 4500 - 4300 yrs BP.  But their own table of dates shows that only five or six dates fall into this period, whereas about twenty do not!  Just as we argued at Rhosyfelin, there is absolutely no reason to connect any of these dates with quarrying activity, and the only way for any reasonable person to interpret them is to say that there was a long history of intermittent occupation by hunters and gatherers -- just as there was at Rhosyfelin.  The authors argue that there was a "slightly longer chronological span" at Rhosyfelin.  Who are they trying to kid?  Their dates show that the site was occupied, off and on, between the Mesolithic and the Iron Age,  with a time span of 6,000 years or more......

There is then an attempt by the authors to match up their perceived "peak quarrying period" with the age of cremated human bone in Aubrey Hole 32, and they repeat the assertion that the holes held bluestones. But as we know, no sound evidence has ever been produced that there was any link at all between the bluestones and the Aubrey Holes.  As ever, we are solidly into the realms of fantasy.

Then to the final section called "From the bluestone quarries to Stonehenge".  This is all speculation and no evidence.  At the outset, the authors refer to "the stone-dragging teams"and the various routes they might have followed.  Then they seek to make a link with Banc Du -- broadly early Neolithic, broadly coincident with the supposed placing of bluestones in the Aubrey Holes, and broadly coinciding with (so they say) the end of megalith quarrying at Carn Goedog.   The trouble is that no connection is established between any of these,  and we could equally well say that everything that happened broadly in the early Neolithic in the British Isles was connected to everything else, as part of a cunning plan.

Banc Du is considered as a possible "stopping place for the bluestones" and as a place where a sort of proto-Stonehenge might have been erected.  But that is considered unlikely, and the latest favourite is Waun Mawn, as we all know, even though Prof Grimes thought it unlikely that there ever was a stone circle here. MPP and his colleagues love the fact that Waun Mawn is only (so they say) 3 km from Carn Goedog, 3km from Rhosyfelin, 2 km from Cerrigmarchogion, and (so they say) 3 km from Banc Du.  A perfect location!  They will soon turn up and will dig it all up.  Watch this space.


My final thoughts about this paper?   Well, I thought that the 2015 paper on Rhosyfelin was one of the worst papers I had ever read.  This one is even worse. It is a mish-mash of assertions, and very little of what the authors say is actually supported by their own cited evidence or by their own radiocarbon dates.

Just a reminder -- this is from the Rhosyfelin dig.  The archaeologists refer to this as "vertical drystone walling and revetment at the edge of the loading platform..."  They apparently expect us to believe this.

More from the Megalithic Quarrymen (3)

Herewith chapter 3 of my scrutiny of the latest paper from the quarrymen, which deals (mostly) with Carn Goedog, which has previously gone unreported in peer-reviewed journals.  Lots of mentions in glossy popular magazines, but after three seasons of digging (2014, 2015 and 206) there have been no field reports or annual updates -- just lots of assertions, photos and diagrams which give the impression, as ever, that everything is sorted and that there are no doubts about anything.........

This is the paper:
"Megalith quarries for Stonehenge’s bluestones", by Mike Parker Pearson, Josh Pollard, Colin Richards, Kate Welham, Chris Casswell, Duncan Schlee, Dave Shaw, Ellen Simmons, Adam Stanford, Richard Bevins & Rob Ixer. Antiquity, June 2018 .

So we come to tools and quarrying techniques.  It's worth quoting this in full:

"...........the only surviving artefacts are of stone. These include a large number of coarse stone tools, a handful of quartz flakes and a single flint blade. The most common of the coarse stone tools are implements with wedge-shaped profiles (n=15), generally with a wide ‘blade’ at one end and a narrow, thick terminal at the other end (Figure 8). These exhibit numerous flake scars along the blade, along one or both sides of the implement. Some also have traces of battering on their thick ends. There are also occasional scars running between the blade and the thick end.
These wedge-sectioned coarse stone tools are interpreted as wedges, for opening up the joints between each naturally formed pillar so that it could be levered away from the outcrop. These tools are all of mudstone or sandstone, not the locally occurring dolerite, which raises the possibility that they were used because of their softness in relation to the spotted dolerite. One reason for this may be that the forces created by driving in soft stone wedges would not have caused fractures within the dolerite that might have weakened the pillar. Any fracturing would have been confined to the soft wedges and not to the monoliths."

Excuse me, gentle reader, but I really am trying to be polite here.  Have I got this right?  The quarrymen are saying that our Neolithic ancestors used soft shale and sandstone "wedges" which were hammered into cracks or joints in the spotted dolerite rockface to assist in the extraction of monoliths, and that the wedges needed to be soft so that they could be sacrificed without causing fractures in the dolerite?  What sort of fantasy world do these people exist in? If you have got metal wedges, that's fine; you might use them in the knowledge that some might be lost if a joint or fracture does not open up as expected, but they might just work.  But we are talking here about the Neolithic, not the Iron Age.  Wooden wedges might work too, if you hammer them in dry and then let them expand as they get wet.  But to use soft and flaky shale as wedges in spotted dolerite bedrock?  No, no, no.........

And what of the evidence that these wedge-shaped bits and pieces are really "implements" or engineering accessories?  Flake scars?  Battering traces?  Scars running from one end to the other?  We need to see the colour of the evidence, and all we have is this -- Figure 8:

Figure 8. Stone wedges and a hammerstone (bottom left) from Neolithic contexts at Carn Goedog (drawn by Irene de Luis).  

Note:  what are "Neolithic contexts"?  That's a nice little throwaway phrase, designed to cement the thesis.  But it is meaningless.  It probably simply means that these bits and pieces were found among the rocks and in the sediments at Carn Goedog.............

As far as I am concerned, bits and pieces of shale, mudstone and sandstone that look like these featured fragments litter the north Pembrokeshire landscape.  I found some lovely elongated and wedge-shaped pieces of shale, complete with facets (or shall we call them flake scars?) in the stream at the bottom of our garden, just the other day.  No attempt is made to show us that these fragments are special in any way -- we are just told that they are.  Not good enough.  In my view, this is just another example of the researchers being desperate to find something that can have been fashioned by human beings -- this will do, they reckon, and we are expected to believe them......

Then there is reference to "a handaxe-like implement with battering along its edges. Such a tool could have been used to open up narrow joints, by widening the sides of the joint sufficiently for a wedge to be then inserted and driven in."  We are not shown this so-called implement, but we are shown (in Figure 9) an apparently widened or broken joint  which could be very old or very new, or anything in between.  It could be natural and it could be man-made -- and we have no idea how unique or unusual it is. The authors tell us that they have seen two such features at Carn Goedog -- but they have not checked to see whether similar features occur on all of the other spotted dolerite tors in this part of Preseli.   As ever, their assumptions of uniqueness and significance are unaccompanied by any studies of control sites with which comparisons could be made. 

There follows a rather fanciful description of how wedges, ropes and timbers were used in the quarrying operations, to facilitate delivery of the delicate monoliths down onto the "platform" which has been referred to earlier.  The idea is that the monoliths would have been levered to the outer edge of the platform and then lowered onto a sledge and hauled away with ropes towards their ultimate destination.  The authors admit that there are no gullies, grooves, prepared surfaces or trackways that might have been used for stone transport.  They suggest that as in the last few centuries, stones would simply have been dragged away across the dry turf surface.

Then the authors say: " The ditch, when open and with its rubble fill, would have blocked movement from the platform and the outcrop, and may have been dug to decommission the quarry. "  (See my previous post for reference to the "ditch". )  This again is fanciful in the extreme.   Are we really expected to believe that Neolithic quarrymen actually decommissioned their putative quarries by digging ditches across their entrances?  Or indeed by filling their ditches with rubble?  Why would they not just have walked away, or moved on to pastures new?

The obsession with storytelling trumps everything.  I cannot understand why the editor of "Antiquity", or the referees, allowed this section (or indeed the whole of this paper) to have been based on the assumption or the ruling hypothesis that this was a Neolithic bluestone quarry, without any consideration whatsoever of the role of natural processes or indeed of the effects of intermittent occupation by people who were simply hunters and gatherers with no interest in quarrying.  I am not averse to the idea that some of the "implements" described may be manufactured;  but why could they not have been scrapers, cutting tools or hammer stones used in the fashioning of other implements required in normal daily life?  This refusal to even contemplate alternative explanations for features is one of the most depressing features of this paper.

Then comes the coup de grace.  Having explained at some length how the platform was used for the movement of monoliths intended for Stonehenge (ie around or earlier than 5,000 years BP) the authors tell us that the buried soil under the platform or within crevices in it contains six pieces of charcoal whose ages range from around 4,500 yrs BP to 4,300 yrs BP.   In other words, they claim that the platform must be later than that, in which case it can have had nothing whatsoever to do with Stonehenge bluestones.

Oh, what a tangled web they weave..........

Wednesday 22 August 2018

More on the Middle Devensian

This is a summary of what we know about the Middle Devensian "warmer episode".  There was probably ice in the uplands of Scotland -- and maybe in Wales as well.  But there was salty tidewater in St George's Channel, and probably no calving ice either.  The extract below relates to Scottish evidence in particular.

What we used to call the "Upton Warren Interstadial" now appears much more complex......,_Quaternary,_Cainozoic_of_north-east_Scotland&action=edit

Middle Devensian (OIS 3)

Periglacial and glacial environments prevailed across much of continental north-west Europe during OIS 3 and glaciers probably existed in the western Highlands for most of the time. There were relatively warm periods between 50 to 41 ka and 37 to 36 ka on the Continent (Huijzer and Vandenberghe, 1998), the former being correlated with the Upton Warren Interglacial of the British chronostratigraphy (P915253; P915254). There is also evidence of two interstadials at roughly equivalent times in the Scandinavian record (P915290). The younger of the two, the ‘Sourlie Interstadial’ is apparently represented by organic deposits beneath till in the lowlands around Glasgow, where reindeer, woolly rhinoceros and mammoth roamed in a tundra-like environment (Jardine et al., 1988; Sutherland and Gordon, 1993). Reindeer bones found in caves near Inchnadamph, Ross-shire, also date from the Middle Devensian (Lawson, 1984). However, there are no known representative organic deposits in north-east Scotland, although the district was probably free of ice. The Crossbrae Peat (Site 5) was originally thought to date from between 22 and 26.5 ka BP, but it is now correlated with OIS 5a or 5c (Whittington et al., 1998).

Sand within the glaciofluvial Byth Gravel at the Howe of Bythe Quarry (Site 6) has yielded luminescence ages of about 45 and about 37 ka, implying the presence of glacier ice in the vicinity (Hall et al., 1995b). This Middle Devensian glaciation would correlate with the Skjonghelleren glaciation of Norway (P915290), when ice probably crossed the North Sea basin (Carr, 1998; Sejrup et al., 2000).

Amino-acid ratios and radiocarbon dates on shells within rafts and tills in the Whitehills Glacigenic Formation suggest that the deposits are derived from cold-water marine muds of Middle Devensian age. More specifically, the ratios correlate with the Bö Interstadial of Norway (P915290), for which ages from 40 to 80 ka have been proposed by Miller et al. (1983) and the higher estimates are favoured by Sejrup et al. (2000). The rafts of the Clava Shelly Clay near Inverness are also thought to have been originally deposited during that interstadial (Merritt, 1992b). Oddly, there appear to be nocorrelatives of the younger Ålesund Interstadial of the Norwegian sequence in north-east Scotland (Peacock and Merritt, 1997).


On the matter of Mid Devensian sea-levels, this is a figure I have published before:

It shows that in the Mid Devensian there was still so much ice on land globally that sea-level was well depressed -- the figure for the interstadial period around 40,000 years ago seems to be c -60m to  -40m.  That gives us water in St Gerge;s Channel -- but the floor of Cardigan Bay might have been dry (or probably soggy) land.  But we still do not know how much isostatic depression there might have been; and we do not know how much Late Devensian glacial and fluvioglacial sediment has accumulated and contributed to the shallowing of the Mid Devensian sea floor.

Mullock Bridge looks important again...

Littorina littorea -- a key indicator of warmer conditions

This paper from 1972 -- yes, it really was that long ago -- starts to look very important again. This was the paper in which I questioned the placing of the Devensian Irish Sea Glacier edge across north Pembrokeshire -- something that was more or less universally accepted at the time.  I argued that the ice edge must have been somewhere near the mouth of Milford Haven, if not further south, because I could only explain the Mullock Bridge kame terrace as an ice edge feature associated with active ice in the late Weichselian / Devensian episode.  It could not, I argued, have been formed earlier, or later.

I had quite an argument at the time about my radiocarbon dating of shell fragments from the "shelly layers" exposed in the gravel pit.  The dates on two samples came out as around 38,000 yrs BP, which was out towards the limit of radiocarbon dating at the time.  Furthermore, argued my critics (including David Bowen, Fred Shotton and Francis Synge), the dates could not be considered reliable because they came from aggregate samples of small shell samples which might have been of many different ages.  The lab (NPL) was pretty sure the dates were good -- and so was I.  And to show a degree of reliability they tested just one species -- Littorina litorrea -- and came up with virtually the same date as the bulk sample dated earlier.  Littorina cannot live where there is much sea ice -- so I assumed, with specialist advice, that this was an indicator of interstadial conditions preceding the glacial episode responsible for the creation of the kame terrace.  So -- a Middle Weichselian interstadial followed by a Late Weichselian glaciation.  Between 50,000 and 30,000 yrs BP, there seem to have been non-glacial conditions in Cardigan Bay, with quite violent temperature oscillations.  Five or six different "interstadial peaks" seem to be present in the temperature record.

After all these years, by dating and my elucidated sequence of events have not been challenged -- indeed, multiple studies by others in the years that have passed have all come to the same conclusion.  So I consider my hypothesis validated or confirmed.

Below I have copied some of the key pieces of information from the PGA paper.


A Late Weichselian kame terrace at Mullock Bridge, Pembrokeshire
by Brian John
December 1972
Proceedings of the Geologists Association 83(2):213-229


A terrace of fluvioglacial sands and gravels is described from Mullock Bridge, near Dale. From its internal characteristics and morphological relationships it is interpreted as a kame terrace formed during the wastage of the Irish Sea glacier at the close of the area's last glaciation. The stratigraphy of the Mullock Bridge deposits is compared with that from four nearby sites, and, following an analysis of derived marine mollusca in the gravels, two radiocarbon age determinations are reported. Each line of evidence suggests that the last glaciation of the Dale area occurred during the Late Weichselian, and that the limit of the Irish Sea glacier lay at least as far south as Milford Haven.

Tuesday 21 August 2018

South Pembrokeshire Devensian till exposures

Devensian till exposures in gullies and on clifftops are so frequent in north Pembrokeshire that there is no great point in pinpointing the localities on a map.  But South Pembrokeshire is a different matter, and for decades it has been assumed that the Devensian ice of the Irish Sea Ice Stream remained somewhere out to the west and did not touch the present coastline.  (Bear in mind that sea-level was around 120 m lower than it is today, and that the coastline was somewhere out towards the shelf edge, far to the SW.)

Well, I have been arguing for years that there was something which we can call the Carmarthen Bay ice lobe, and that the ice did indeed press as far east as Caldey Island -- and that if it did that it must also have touched the south coast cliffline in multiple locations.  I didn't have much evidence to go on, apart from some suggestive comments in the old Geological Survey memoirs written by Dixon, Cantrill and others.  Anyway, in the spring of this year I gathered evidence to supplement my earlier observations at places like Westdale, Mullock Bridge, West Angle, and Ballum's Bay on Caldey. Some of the earlier evidence is here:

There is another apparent till deposit at Freshwater West, exposed after freak storm conditions, which I have been unable to examine myself:

You can find other posts by using the search facility on this blog.

In the spring of this year, in the course of three walks along the south Pembrokeshire cliffs (in the company of wife and several friends, who were in much more of a hurry than I was) I found a number of localities where unconsolidated till can be seen on clifftops and in gullies.  I have recorded my observations and photos here:

It's interesting that I have walked these cliffs before, many times, without properly observing or recording these occurrences -- which only goes to show that once an area is researched by somebody, that is not the end of it.  It is very easy to miss things....... and maybe, as you become more experienced and know what you are looking for, you see things that you have previously looked at without realising their significance............

There are amazing similarities between these south Pembrokeshire till occurrences and those in the Scilly Isles.  I will develop this in another post.


The upshot of all of this is that what started in my mind as a working hypothesis is now -- in my mind at least -- a pretty reliable confirmation that the ice of the Devensian ISIS was in contact with the south Pembrokeshire cliffline at least as far east as Caldey Island.  And it's gratifying to see that the field workers of the BRITICE / CHRONO project are accepting this too.  I must try and convince some of those guys to come over and take some samples for cosmogenic dating,  just to add some scientific weight to my simple glacial geomorphology!

Monday 20 August 2018

Topographic controls over ice stream behaviour

They are coming thick and fast.  Yet another paper relating to the deglacial phase of the Devensian Irish Sea Ice Stream.  This is a very technical paper, somewhat difficult to penetrate, but again there is very useful information in it.  Underpinning the conclusions of the authors are more luminescence and cosmogenic age determinations, particularly from the SE coast of Ireland.  I remember visiting those sites almost 60 years ago in the company of Francis Synge and Frank Mitchell..........

The new work is very relevant to West Wales because whatever went on here was mirrored pretty closely on the other side of St George's Channel.  So the sequence of glaciation and deglaciation over there will give us more than a few guidelines......

The model now looks like this:  Maximum extent into the Celtic Sea around 27,000n years ago.  Then extremely rapid (300–600 m per year) retreat from maximum extent, a slowing of retreat (26 m per year) during the period 25.9–24.2 ka, ice margin stabilization (3 m per year; 24.2–22.1 ka), rapid retreat (152 m per year; 22.1–21.6 ka), and finally a return to slower retreat rates (21 m per year; 21.6–19.5 ka).  The right-hand diagram reproduced above shows the assumed ice edge positions looping across St George's Channel towards Pembrokeshire in phases 1-7, each one tied to a specific Irish location.  Quote:  "Our Bayesian age model indicates that initial ice marginal retreat onto the southern Irish coast occurred at 25.9 ± 1.4 ka (Boundary 2). Retreat of the ISIS from the southern coast of Ireland is constrained by the modeled age (Boundary 3) of 25.1 ± 1.2 ka. Deglaciation to the Wexford coast, and associated deposition of the Screen Hills complex, occurred between 24.2 ± 1.2 ka and 22.1 ± 0.7 ka (Boundaries 4–7)."   So this must have been the critical period for the wastage of ice from western Pembrokeshire.  Let's hope that dating work is planned for some of the key Pembrokeshire sites...........

Now I'm going to have another gripe about the flowlines on the left-hand and middle maps.  The easternmost flowlines in the area under scrutiny cannot be correct; and it is ironic that in a paper devoted to topographic controls and trough geometry the authors are showing ice flowing from NNE towards SSW where there appears to be no topographic control whatsoever, and where the laws of ice physics say that flow must have been from NW towards SE,or perpendicular to the ice edge.  I will keep on banging on about this until somebody shows me some evidence that I am wrong.  Heyho -- all good fun!

On balance, another great contribution from members of the BRITICE-CHRONO research team.

Trough geometry was a greater influence than climate-ocean forcing in regulating retreat of the marine-based Irish-Sea Ice Stream (2018)
by David Small, Rachel K. Smedley, Richard C. Chiverrell, James D. Scourse, Colm Ó Cofaigh, Geoff A.T. Duller, Stephen McCarron, Matthew J. Burke, David J.A. Evans, Derek Fabel, Delia M. Gheorghiu, Geoff S.P. Thomas, Sheng Xu, Chris D. Clark

GSA Bulletin (2018) May 28, 2018

Marine terminating ice streams are a major component of contemporary ice sheets and are likely to have a fundamental influence on their future evolution and concomitant contribution to sea-level rise. To accurately predict this evolution requires that modern day observations can be placed into a longer-term context and that numerical ice sheet models used for making predictions are validated against known evolution of former ice masses. New geochronological data document a stepped retreat of the paleo−Irish Sea Ice Stream from its Last Glacial Maximum limits, constraining changes in the time-averaged retreat rates between well-defined ice marginal positions. The timing and pace of this retreat is compatible with the sediment-landform record and suggests that ice marginal retreat was primarily conditioned by trough geometry and that its pacing was independent of ocean-climate forcing. We present and integrate new luminescence and cosmogenic exposure ages in a spatial Bayesian sequence model for a north-south (173km) transect of the largest marine-terminating ice stream draining the last British−Irish Ice Sheet. From the south and east coasts of Ireland, initial rates of ice margin retreat were as high as 300−600 m a−1, but retreat slowed to 26 m a−1 as the ice stream became topographically constricted within St George’s Channel, a sea channel between Ireland to the west and Great Britain to the east, and then stabilized (retreating at only 3 m a−1) at the narrowest point of the trough during the climatic warming of Greenland Interstadial 2 (GI-2: 23.3−22.9 ka). Later retreat across a normal bed-slope during the cooler conditions of Greenland Stadial 2 was unexpectedly rapid (152 m a−1). We demonstrate that trough geometry had a profound influence on ice margin retreat and suggest that the final rapid retreat was conditioned by ice sheet drawdown (dynamic thinning) during stabilization at the trough constriction, which was exacerbated by increased calving due to warmer ocean waters during GI-2.


The geochronological data presented here allow us to test a conceptual model of ISIS deglaciation in south and east Ireland inferred from the sediment-landform assemblage record. Integration of new geochronological data using Bayesian age modeling produces a conformable age model that supports the conceptual model of deglaciation, with extremely rapid (300–600 m a–1) retreat from maximum extent, a slowing of retreat (26 m a–1) during the period 25.9–24.2 ka, ice margin stabilization (3 m a–1; 24.2–22.1 ka), rapid retreat (152 m a–1; 22.1–21.6 ka), and finally a return to slower retreat rates (21 m a–1; 21.6–19.5 ka).

This timescale strongly suggests that aspects of ISIS behavior during deglaciation displayed a complex relationship to external climatic forcing. Extremely rapid advance of the ISIS to its maximum extent, and its subsequent retreat at 26–25 ka is not directly correlated with distinct climate forcing in the North Atlantic region. Such behavior may be explained as a dynamic instability in response to overextension of the ice stream to the maximum limit that rendered it vulnerable to rapid retreat. Similarly, the stabilization of the ice margin at the Screen Hills spans a time of distinct warming (GI-2) with the subsequent rapid retreat occurring during colder conditions of GS-2. The stabilization at the Screen Hills is the most distinct change in pace of ISIS retreat evidenced by the data presented here and it occurs where there is a step-change in the con ning trough geometry highlighting the important role that this plays in condition- ing ice margin retreat. However, contrary to this, the ISIS subsequently underwent rapid retreat without major changes in trough geometry. We speculate that this represents a delayed response of the ice margin to the climate forcing of GI-2. Overall, changing trough geometry and internal feedbacks related to the overextension, retreat, and stabilization of the ISIS appear to obscure the role of external drivers such as climatic forcing.

The conceptual model and geochronological data presented here provide evidence for specific ice margin behavior during overall deglaciation that provides a testing ground for numerical models that likely require high resolution representations of grounding line dynamics. As contemporary ice streams in Greenland and Antarctica evolve in response to anthropogenic climate change they will undergo retreat that is conditioned both by climatic forcing and their internal dynamics. Our data highlight the potential for the evolution of rapid ice margin retreat to be highly nonlinear and conditioned strongly by trough geometry.