Kaldalon in its regional context

Vestfirdir -- almost an island, where the growth and retreat of glaciers show some connections with the East Greenland coastal zone and some with the Iceland mainland.  Here the Dranga and Glama plateaux have both supported ice caps that have come and gone, sometimes in sync with the ice masses elsewhere, and sometimes not.......

The Dranga and Glama ice caps, according to various authors.   At intervals following the collapse of the LGM ice cap over Vestfirdir, these two ice caps have coalesced into a single unit.

Landscape types in Vestfirdir.  At times the plateau supported a singe ice cap with two m ain cemntres of accumulation and radial ice flow.

I have already flagged up the connections between East Greenland and NW Iceland, but as research proceeds, the wider the scatter of dates there seems to be, for the key glacial events such as glacier extinctions, rejuvenations and surges.  Both Drangajokull and Glamujokull seem to have melted away completely during the Holocene.  Both seem to have been regenerated during the Little Ice Age, but subsequently Glamujokull has disappeared again while Drangajokull has survived.  The three outlet glaciers of Drangajokull -- namely Kaldalonsjokul, Reykjafjardarjokul and Leirufjardarjokul -- have all surged intermittently within the past 300 years, but not entirely in harmony.

These are the dates of the recent (Little Ice age) surges:

Reykjarfjarðarjökull 1846, ~1886, ~1910, 1934–1939, and 2002–2006
Leirufjarðarjökull ~ 1700?, 1846, 1898, 1938–1942, and 1996–2001
Kaldalónsjökull ~1740, 1780?, 1810?, 1820, 1860,1920, 1936–1940, and 1996–2001

These surges are controlled by regional climatic variations, the precise glaciological characteristics of the glacier catchment areas, and by volcanic events and ash clouds (layers of ash on glacier surfaces can have a significant effect in lowering albedo and increasing melting rates.

Ice edge positions following the LGM and preceding the Little Ice Age are difficult to establish with certainty.  The assumption that the big moraines in the Reykjafordur and Kaldalon valleys are of Younger Dryas age are looking a little shaky -- and in now seems (from recent work) that they may have been formed a little later, in the Pre-Boreal period, around 10,300 - 9,000 yrs BP.  They are related to the moraines of the Budi stage in South Iceland, but are not exactly synchronised.

When the outermost moraine in Reykjafjordur was formed, relative sea level seems to have been around 32m asl.  The moraine has a planed top, and there are exposures of washed rock surfaces and washed till in the vicinity.  Was this the equivalent of the Alftanes stage (Older Dryas) in south Iceland?  Quite possibly.  And does the Seleyri moraine in Kaldalon date from the same stage?  That seems logical to me..........

This map of glacier stages in Reykjafjordur is based on fieldwork by the DUVP team in 1975.  Reference -- John 1975.   It shows three major ice front positions in the lower part of the valley.  The more recent research by Brynjolfsson and others concentrates on the moraines in the upper valley which are related to the Little Ice Age.

In this glacial chronology for Reykjafjordur two groups of moraines are identified.  Moraines M1-M4 date fromn the Little Ice Age.  Moraines O0 - O4 are older, dating from the Late Glacial and Neoglacial glacial episodes.  Ref:  John, 1975

European Glacial Landscapes
The Holocene
2024, Pages 193-224
Chapter 12 - Holocene glacial history and landforms of Iceland
Ívar Örn Benediktsson, Skafti Brynjólfsson, Lovísa Ásbjörnsdóttir, Wesley R. Farnsworth 

https://www.sciencedirect.com/science/chapter/edited-volume/abs/pii/B978032399712600012X?via%3Dihub



Extract from Abstract

Climate amelioration in the Early Holocene was punctuated by the preboreal readvance of the Iceland Ice Sheet (IIS) between c. 11.3 and 10.4 cal ka BP, as manifested by raised shorelines and ice-marginal landforms below and inside Younger Dryas shorelines and moraines, respectively. By about 9 ka BP, most glaciers had become smaller than present and crustal rebound was completed. The Early Holocene retreat of the IIS is widely marked by end moraines in the highlands. Declining temperatures after the Holocene Thermal Maximum (HTM c. 8.5–6.5 cal ka BP in Iceland), during which glaciers were largely absent, resulted in the onset of the neoglaciation and the reformation of ice caps at 5.5–4.4 cal ka BP, with many outlet glaciers reaching their maximum Neoglacial extent at different times until as late as 1.1 cal ka BP. The Little Ice Age maximum glacier extent occurred either in the 1700s or the late 1800s. Since then, Icelandic glaciers have been retreating apart from observed still-stands or readvances from the 1960s–70s until the mid-1990s.

================

In 1983 Eggert Larusson, who was a team member of our Durham University Vestfirdir Project in 1973-77, tried to unravel the sequence of glacial events in his 1983 doctorate thesis. he also tried to relate these stages to marine limits and lower terraces and other shorelines. He referred to "Latragrunn stage" when Vestfirdir was completely covered by its own independent ice cap --- during the LGM. Then there was a "Tjaldanes stage" when relative sea-level was around 11-22m asl -- and this must be the equivalent of of the "Jokulgardir stage" in Kaldalon and the Kirkjubol stage in Reykjafjordur. http://etheses.dur.ac.uk/7787/

Larusson, Eggert (1983) Aspects of the glacial geomorphology of the Vestfirdir Peninsula of northwest Iceland with particular reference to the Vestur-Isafjardarsysla area.
Doctoral thesis, Durham University.

Abstract
The evolution of the landscape of Vestfir6ir, made almost entirely of volcanic rocks, is traced from the lilocene, when the oldest rocks formed, through the Pliocene and Pleistocene. Volcanic activity ceased first in the north western part leaving a basalt plateau with occasional large volcanoes protruding. Fluvial erosion, guided by a westerly dip of the plateau and tectonic lineaments, left a well developed drainage pattern there by the rime volcanic activity ceased in the southeast. The snowline fluctuated widely during the Plio-Pleistocene. Cirque and valley glaciations were very effective in sculpturing the landscape where the preglacial relief was greatest, in the northwest. Ice sheet glaciations affected the whole peninsula and offshore areas with linear erosion dominant in the northwest and areal scouring elsewhere. The glacial geomorphology of Dyrafjorour and northern Arnarfjorour is mapped. The highest marine limit is in the Nupur area, about 110 m, and shorelines and marine limits higher than 70 m are at 7 other localities at least. At least' two stages of glacial readvances are recognized: The Tjaldanes stage occurred when sea level was between 11 and 22 m and is probably of "Younger Dryas" age; later a readvance occurred in the cirques in the area. On the basis of evidence on cirque distribution, cirque elevation, zeolite zonation, distribution of glacial erosional landscapes, glacial history, marine limits, ice cap profiles and shelf moraine a model of maximum glaciations of Vestfir6ir is proposed: The whole of Vestfirdir and the surrounding shelf areas was completely ice covered with no ice free areas. Such a stage of glaciation, the Latragrunn stage, probably prevailed in the Vestfirdir area during the last glaciation.


=====================


See also:


https://www.researchgate.net/publication/377874262_Moraine_stages_in_the_Reykjarfjordur_outlet_glacier_trough

The big moraines of East Greenland and Vestfirdir

Winter satellite image of the Holger Danskes Briller terminal moraine in Kjove Land, East Greenland. This is referred to as a delta moraine, because it is associated on its outer edge with a fluvioglacial terrace at 101m asl.  The lake inside the moraine is to the left of the ridge.

The largest terminal moraine in Vestfirdir, NW Iceland -- in the glacial trough of Kaldalon.  Summer satellite image.  The up-glacier side is to the right.  Sandur and tidal flats to the left, with gravel track and landing strip.  

It is tempting to correlate the big terminal moraine that blocks the eastern end of the Holger Danskes Briller trough in Kjove Land (East Greenland) with the big terminal moraine in Kaldalon which is labelled as Jokulgardur.  They are both significant landscape features, and both represent the stillstands of significant glaciers within glacial outlet troughs.  Well, East Greenland is not so far away from NW Iceland, and the two regions must have experienced broadly similar climatic shifts following the LGM glacial episode.  There are differences, of course; the glaciers associated with Drangajokull are much smaller than those associated with the ice cap in East Greenland, and so might be expected to react more quickly to climatic oscillations.  The histories of relative sea level are very different on both sides of the Denmark Strait.  But there was surging glacier behaviour in East Greenland just as there was in Vestfirdir..........

The Milne Land stage in Greenland is the approximate geological time equivalent of the Búði stage in Iceland. Both represent significant glacial events that occurred during the Late-Glacial period, specifically correlated with the Younger Dryas climatic "event" or possibly a few centuries later.

Milne Land stage (Greenland): This stage refers to a period of significant glacier advance or stillstand in East Greenland, marked by prominent moraine systems. It is primarily dated to the Younger Dryas chronozone, approximately 12,800 to 11,500 years ago (or slightly before this time, in the Pre-Boreal).

Búði stage (Iceland): This stage also represents a major readvance of the ice sheet in Iceland. It is generally correlated with the Younger Dryas cold period, occurring around 11,000 to 10,000 years BP, although some authors now suggest it formed later, during the Pre-Boreal period.

Both stages reflect regional climatic shifts near the end of the last ice age, making them correlative in geological time within the North Atlantic region.

Kaldalon -- the 1820 surge?

 

Traces of the 1820 (?) morainic loop on the floor of the Kaldalon Valley.  These are in the form of low elongated ridges and hummocks, quite subtle but nonetheless noticeable when one walks along on the valley floor towards the glacier.  The kame terrace remnants on the valley side above must be unrelated.  They may be related to the 1780 surge for which no terminal moraine is known (it may have been buried beneath the accumulating sandur).

The photo below (taken from high on the valley side) shows the degraded terminal morainic ridge with a couple of small "inliers" in the sandur, surrounded by gravel and water, further out.

The Trimbilsstadir double moraine

The Kaldalon "valley side esker" (now interpreted as a moraine) seen from the north.  The steep ice contact face is prominent, facing up-valley.  The older, subdued moraine is seen to the right, in contact with the outer slope..  

Satellite image of the Trimbilsstadur moraine / kame complex.  the outer (older) ridge is subdued and well vegetated.  The inner, very prominent, ridge is not grassed over to the same extent, and has a steep up-valley face interpreted as an ice contact slope.  The associated linear kame terraces (or lateral moraines!) are seen right of centre.

Lateral morainic ridge connected to the inner Trimbilsstadir end moraine, which runs steeply downslope.  We see the ice contact face. (View down-valley, with the Younger Dryas terminal moraine in the middle distance.)

This troublesome and confusing landform has been the subject of much debate in the literature, and I think that our 1962 designation of the large curved ridge (on the inside or eastern flank of the subdued green moraine)  as a "valley side esker" was somewhat naive and premature.  Others who have examined the feature more recently seem to be convinced that it is made primarily of till, not fluvioglacial sands and gravels, and that is important information.

David Sugden and I always were a bit worried about an apparent "esker" being formed this high on a valley slope -- especially since eskers always form in tunnels with ice above and on either side of them.  The eskers on the floor of the valley do seem to have been formed in sinuous tunnels beneath a wasting glacier, and they are linked into assorted hummocky dead-ice features and also to an abandoned outwash fan.  But here on the hillside it is difficult to conceive of "containing"  ice on the down-valley flank.  Neither do the conditions exist for classic kame terrace formation, in contrast to other situations further up the valley........

So I now revise my opinion, and think that this feature is a remnant of a terminal moraine with a very sharply defined up-valley ice contact face.  The feature is also much fresher than the subdued well vegetated moraine on its outer edge -- so it has to be younger.

I think we can also say something about the nature of the glacier advance associated with the ridge.  Because there is a substantial "void" on the up-glacier flank, this suggests to me a very rapid and dramatic ice advance involving relatively clean ice.  (If the advance had been more gradual and prolonged, the ice would have been heavily laden with supraglacial and englacial debris, and this load, when dumped, would have left a dramatic expression in the landscape.)

So going with the theory that this moraine was formed around 1740, what do we know about this advance?  Now we would call it a surge.   The old records refer to a great readvance between 1700 and 1756, in which the glacier covered once green meadows and destroyed the farm of Trimbilsstadir.  The ice front moved forward by at least 2 km.  There are also records of a great "glacier burst" or jokulhlaup in Kaldalon in 1741.  This suggests massive quantities of meltwater associated with a catastrophic ice wastage event.

I think we have no option but to argue that there were TWO surge advances here, ending up in virtually the same place.  Coincidence?  Or was there cause and effect?   Of course, it is quite possible that the older moraine was once more extensive, and that it did act as something oif a barrier to the advance that occurred around 1740.

Ages?  From the evidence presented by other workers, it is most likely that the old green moraine is of Neoglacial age, maybe dating to 3500 - 2500 yrs BP, and that the newer moraine segment dates from the Little Ice Age surge that occurred around 1740.

The "matching feature" on the north side of the valley, which I have referred to as the Kegsir Moraine, is even more copmplex, with incorporated distorted peats and other layers demonstrating a bulldozing effect during one or both of the advance phases.

How "Kaldalon Kame Terraces" are formed

The "Kaldalon kame terraces" at the base of the Votubjorg cliff, on the south side of the glacial trough.  In places there are clear trenches or gullies behind them, cut against bedrock exposures. 

They are very prominent features.........

The landforms which are most difficult to explain in the Kaldalon Valley in NW Iceland are the elongated features on the valley sides that are referred to variously by assorted researchers as "lateral moraines" or as "kame terraces".  Some refer ro them as being made predominantly of fluvioglacial sands and gravels, while others say they are made predominantly of till or moraine, and others say that they are essentially made of scree or rockfall debris.  Actually we are all right, because the characteristics of these ridges vary, depending on where you happen to look or where you dig your hole in the ground.

I have been examining some of the archive photos from 1960 and 1974, and from the work we did around the glacier snout one can see what the key processes and development stages are.

Stage 1.  On this satellite image you can see that the meltwater river is tight up against the western rock wall of the trough.  Here, because of enhanced ablation or melting, the glacier cross profile is convex.  This means there is a trough or gully along the glacier edge, into which surface meltwater is channelled.  This meltwater comes from surface streams on the bedrock on the glacier flanks or from surface glacier streams from higher up the glacier.

Stage 2.  As melting proceeds, the surface meltwater streams disappearinto the glacier, and most high-volume flow occurs sub-glacially, in tunnels beneath the ice but still against the rock wall.   View towards the rock wall, showing cascades of meltwater flowing down towards  and then into the glacier.

One of the subglacial drainage routes, now abandoned.

Stage 3.  The ice marginal gully is still visible, but it is dry because all meltwater has now been diverted beneath the ice.  Surface morainic debris (including till and faceted, abraded and striated boulders) begins to accumulate in the abandoned stream gully.  There are additional inputs from rockfall debris and scree cones derived from the high overlooking basalt cliffs.

Simultaneously, fluvioglacial materials accumulate in the main stream discharge routes, within and along the edge of the glacier.

Stage 4.  Deglaciation and glacier retreat leaves high marginal terraces on the valley side composed of fluvioglacial materials below and capped withy a mixture of till and rockfall debris.  This association is NOT the resuklt of ice front oscillations, but simply a consequence of the meltout process. The ridge is broken into segments, but overall the gradient is down towards a past snout position on the valley floor.

PS. Eskers are often capped with morainic debris in open landscape situations, and with rockfall debris in constrained valley situations.  Here we have both.

PPS.  Interesting question  -- why do these features develop in the glacier on the valley side rather than on the valley floor?  One might think that the main meltwater conduits would form right at the base of the glacier.  I suspect that this has something to do with modest meltwater flows and the tendency for very deep conduits to close under ice pressure during the 8 months or so of freezing conditions.  Shallow englacial conduits against the valley rockwall have a much greater chance of staying open, where there are crevasses and water discharge supplements from spring and autumn meltwater streams flowing down the valley side.

Kaldalon -- the big moraine and the Armuli raised marine terrace

 

The Jokulgardur terminal moraine, generally thought to be of Younger Dryas age, c 11,000 yrs BP.  It is by far the most prominent feature on the floor of the Kaldalon Valley. Marine sediments are exposed on its outer (western) edge.

In the latter part of this long article, Bout et al (1955) discuss the relationship between changing ice volumes and the raised marine terraces of Isafjardardjup.  But they do not seem to make a direct link between the big moraine in Kaldalon and the prominent marine tarrace north of Armuli, at the mouth of the Kaldalon trough.  That terrace, referred to as the 15 - 30m terrace by Bout et al, is the same terrace as that described by my colleagues and myself in 1975 as the 14 - 24m terrace. We need to be flexible in our labelling, because there are substantial variations in the altitude of the terrace top, and its edges are in places difficult to define.........

I am now rather convinced that this terrace represents a marine transgression or stillstand that coincided with the Younger Dryas ice expansion in Vestfirdir.  The is referred to as the "Younger Dryas glacial event" -- which may actually have been two events, incorporating the so-called Budi Moraine in south Iceland and other near-contemporaneous features.  Although global sea level was at c -60m at this time, there was also a contemporaneous very great isostatic load on NW Iceland, associated with a detached Vestfirdir ice cap which incorporated both Glamajokull and Drangajokull.  .  Isostatic uplift might have been slowed for maybe a thousand years, with the rate of uplift then approximately equivalent to the rate of eustatic sea-level rise.

Post-glacial (Holocene) sea-level curve, showing the stillstand or "step" near -60m around 11,000 - 12,000 years ago.  This was associated with the Younger Dryas "glacial event".

As David Sugden and I pointed out in 1962, there are marine sediments (including varved clays and bedded sands and silts) on the outer edge of the big terminal moraine called Jokulgardur or Holar.  These features have been missed or ignored by some other workers.  The sandur level adjacent to the western edge of the terminal moraine varies between 12m and 9m asl. The moraine top varies in altitude between 12m and 27m.  The surface is composed of unwashed till with a scatter of striated erratic boulders.  As far as we are aware, there are no marine terraces (with or without marine mollusca) on the inside of the moraine, even though the dry sandur level is at c 13m.  Therefore it is entirely logical to propose that at the time of the formation of the Armuli terrace glacier ice was blocking the valley as far west as the position of the moraine.  This in turn points to a grounded or floating ice edge.

https://www.researchgate.net/publication/262414815_The_Morphology_of_Kaldalon_a_Recently_Deglaciated_Valley_in_Iceland

The big terminal moraine in profile, seen from the southern gap.  Marine sediments are exposed in places on the steep west-facing slope.

Map of the moraine and features immediately up-valley.  From the DUVP 1974 Field Report.

Sketch map from Bout et al (1955) showing some features in Kaldalon and Skjaldfannardalur to the south.  In the latter valley there are abundant dead ice features linked to a delta near the valley entrance.  In 1976 we found traces of a high sea-level at c 30m on the SE flank of Steindorsfell.  We also estimated that at the time of rapid valley glacier ice wastage, relative sea level here was at 17.6m.  

Géomorphologie et Glaciologie en Islande centrale
 Pierre Bout, Jean Corbel, Max Derruau,  L. Garavel,  Charles-Pierre Péguy:
Norois  1955, vol  8,  pp. 461-574

https://www.persee.fr/doc/noroi_0029-182x_1955_num_8_1_1102

In Kaldalon the outermost moraine on the south shore is near the Kalda stream cutting, just to the west of the Seleyri spit (John and Sugden, 1962). Here there is a distinctive vegetation-free mound of till and striated boulders.  Moraine teraces extend for at least 500m along the valley side.  We speculated in 1962 that this might be a remnant of a terminal moraine, laid down on a grounded ice front at a time of relatively high sea level.  Could this be related to the 30m beach traces on Steindorsfell?

It's interesting that according to Hansom and Briggs (1991) the highest shoreline traces on Hornstrandir, 30 km to the north, are around 26m asl (Hjort et al, 1985).  We defined the upper marine limit as 30m in Skaldfannardalur -- and this is the same altitude recognised by Principato and Geirsdottir in 2002 and based on their own evidence.

Principato (2008) does not recognize the presence of the raised marine terrace at 24m - 14m  near Armuli, referring only to fragments of the very low terrace (which is referred to elsewhere as the "Nucella" terrace) and the marine limit near 30m.  I agree that neither of these can be "tied in" to the prominent moraines in Kaldalon.  But as mentioned above, I am convinced that the massive Jokulgardur moraine was formed at the same time as the 24-14m terrace.  A short-lived marine stillstand and a contemporaneous grounded ice front in Kaldalon -- that makes sense.  The delta moraine on the flank of Steindorsfell was formed at the same time, when the valleys of Skjaldfallardalur and Hraundalur were filled with ice.

My field map showing the main depositional features of the Armuli - Melgraseyri area.

The 24-14m terrace neat Melgraseyri, with the entrance to the Kaldalon trough beyond.  The "Nucella terrace" remnants are seen just above the present shoreline.

Ice front stillstands in the Kaldalon Outlet Trough

 

Top image: Annotated satellite image showing the main geomorphological features of the middle section of the Kaldalon valley.  Based on mapping and field notes from many visits in 1960 and 1973-77.  The literature is filled with references to multiple ice front stillstands, with each one now assumed to coincide with a glacier surge. 

Sketch map:  suggested ice front positions.  The 1780 position is  interpolated from historical records; The 1740 and 1820 positions are based on the mapping of terminal moraine remnants.

There is much accumulated evidence to show that the massive moraine which is the most prominent feature of the depositional landscape is of Younger Dryas age, around 12,000 years BP.  All of the ice front positions recorded on the up-valley side of that moraine must be younger.  It would be tempting to refer to thes as "recessional moraines", but that would be inadequate, since Kaldalonsjökull is one of a group of Drangajökull surging glaciers with a history of spectacular advances of up to 1 km.  Not all of these advances are recorded in the landscape, since now and then large advances might have wiped out the traces of prior (less extensive) retreats and advances........

According to Brynjolfsson et al, there are traces of seven surges, as shown on this image:

There is a certain amount of guesswork involved, because not all advances are recorded by strong linear features such as terminal and lateral moraines, and the valley floor is continuously reorganized -- and raised -- by the shifting sandur of the Morilla River.  There may be many buried moraine fragments related to the five morainic mounds which we referred to in 1960 as "the Green Mounds". 

The MOR 3 morainic ridge (assumed to date from c 1820) was described in detail in the article by David Sugden and myself in 1962, on p 359.  We described abundant fluvioglacial sands and gravels resting on top of morainic material with large striated boulders.  

There have also been catastrophic events:  Quote:  Moraines 4, 5 and 6, described below, are not recognizable  anymore, because of a 1998 outburst flood in the glacial river Mórilla, which drains Kaldalónsjökull (Sigur›sson, 2000). The river transported icebergs of many metres in diameter 2–3 km 

down-valley and deposited an 8- to 10-m thick sheet of coarsegrained gravel on the valley floor distal to the glacier margin. A thinner sheet of gravel was deposited to about 1 km down-valley.

Ice edge positions MOR 4, 5, 6, and 7 are based largely on literary sources, and are approximate only.

The 1740 surge moved the ice front down-valley by at least 1.5 km, and the 1994-1999 surge is known from detailed records to have moved the ice front downvalley by around 1 km.  In between these in the timeline, there were five known surges which involved advances of between 300m and 500m.  

My own map of possible surges and ice edge positions is below:

Were there Neoglacial advances?

There is considerable discussion in the literature about the events of the Neoglacial (the period from about 3,500 yrs BP to about 2,000 yrs BP, during the Holocene.  The labelling is not very satisfactory, since in many areas this period (following the Hypsithermal or Climatic Optimum period, 8,000 - 3,500 yrs BP) was marked by the persistence of relatively warm conditions.  During the Hypsithermal episode many smaller ice masses wasted away completely, whilst in other areas there was renewed glacial activity as a consequence of localised climatic shifts and increases in precipitation.

AI generated chart showing climate oscillations in the Atlantic arena

https://doi.org/10.1177/0959683615576232

Principato and others suggest that the Neoglacial advance of the Drangajökull glaciers was greater than any of the advances associated with Little Ice Age surges -- and the evidence from Leirufjördur and ~Reykjafjördur seems persuasive.  In the Kaldalon trough the evidence is more difficult to interpret, and hinges partly on the interpretation of contorted (ice pushed?) peat beds and other organic layers exposed on the flanks of the Kegsir and Trimbilsstadir moraines.

Stream cutting on the outer edge of the Kegsir Moraine, excavated and described by the members of the Durham University Vestfirdir Project, 1973-77.  Here there are distorted, ice-pushed organic sediment layers including peat beds.

 

We studied the Kegsir moraine in 1975 and 1976, and obtained the following radiocarbon dates from wood fragments contained within contorted peaty layers:

Section A -- depth 295 cm. Wood fragments in basal gravel and stones, 7115 +/- 125 yrs BP (St 5634)

Section A -- depth 260 cm.  Wood fragments in peaty and stony grey silt, 3535 +/- 125 yrs BP (St 5634)

Section C -- depth c 315 cm.  Wood fragments in grey stony gravel, 4170 +/- 95 yrs (St 6051)

Section B -- depth 35 cm.  Wood fragmnents in peat above stony gravelly layer, 255 +/- 95 yrs (St 6050)

These radiocarbon dating results should be interpreted alongside these as reported by Principato.  Samples taken from an apparently undisturbed peaty layer not far from our sampling site gave dates of 3328 +/- 45 yrs BP,  2503 +/- 59 yrs BP and  2623 m+/- 68 yrs BP from wood fragments and mossy materials beneath an ash layer which was itself covered by thin layers of sands and gravels.  Another sample from an unidentified  peat layer is dated at 1296 +/- 39 yrs BP  (Table 1 from Principato 2008).  Might the ash layer date from the Hekla 1104 eruption or one of the others known in the record?  A sample taken from the Trimbilsstadir moraine on the south side of the valley does suggest this association.

Principato (2008) claims that tephra layers found in the "green moraine" on the north side of Kaldalon point to a glacial readvance event in Kaldalon at around 2600-3000 yrs BP.  But her radiocarbion dates appear to contradict this, suggesting that this was a time of peat and tree / shrub growth............  There is a dating dilemma here, which will no doubt be resolved in future work.

Principato claims that there was at least one Neoglacial advance in Kaldalon which was more extensive than any of the Little Ice Age advances.  That may be true, but there are just a few metres in it.  The "double moraine" at Trimbilsstadir shows two readvances to almost the same position. 

At face value, the radiocarbon evidence from the Kegsir Moraine shows that there was tree or shrub growth in the valley from c 7000 yrs BP to c 3500 yrs BP --  and at some stage after that, there was an ice advance which disrupted the organic sediments and possibly resulted in the formation of a push moraine.  Other evidence suggests a date of around 3000 - 2600 yrs BP for this event in Kaldalon -- and that seems reasonable.  One event or several?  The jury is still out.

The Kegsir Moraine and the Trimbilsstadir moraine are both complex features which have attracted much discussion.  I think it is quite possible that the Kaldalon glacier has advanced to this position on two -- or possibly three -- separate occasions, first in the Neoglacial and then again in the Little Ice Age, in 1740 and possibly 1780.

================

The outermost Kaldalon terminal moraine

The big Younger Dryas moraine (Jokulgardur) is not the oldest or outermost moraine in the glacial trough.  There are remnants of a much earlier moraine on the south shore near the trough exit, just to the west of the Seleyri spit.  There is a substantial mound of moraine above the road, and till is exposed in the cutting of the Kolda stream.  This is over 4 km beyond the Jokulgardur moraine.

Exposure of the outermost moraine in the Kaldalon trough.  The Seleyri spit can be seen in the distance.

Nothing is known about thye age of the moraine, but it must be Late Glacial, formed either during a readvance prior to the Allerod Interstadial, or else a retreat stage during the earlier deglaciation of the Vestfirdir Peninsula, at a time of higher sea level.

The boulders that litter the tidal mud flats in Kaldalon -- or at least some of them -- might be ralated to this outermost glacial stillstand stage.....

Shrinkages and surges of Drangajökull since 1850

This is a very useful publication.  Note that a very large shrinkage of the ice cap is shown for the southern sector, and for a large dome on the western flank, between Kaldalon and Leirufjördur.  Recent studies suggest that these areas may not have supported moving glacier ice after 1850 -- it is suggested that there were simply extensive snowfields and firn accumulations which were never thick enough to be transformed into glacier ice. 

I see a possible parallel here in the UK, during the Devensian, with connected snowfields between actual ice masses in small ice caps on Dartmoor, Exmoor, Bodmin, Mendip etc..........

 https://jokull.jorfi.is/articles/jokull2020.70/jokull2020.70.001.pdf

============================

It is now also widely accepted that all three of the main outlet glaciers of Drangajökull are subject to surging behaviour, somewhat out of sync with the other (larger) ice masses of Iceland.  Each surge appears to last for up to ten years -- significantly longer than the surges of the larger glaciers of Iceland.

See also:

Brynjólfsson, S., A. Schomacker, E. R. Guðmundsdóttir and Ó. Ingólfsson. 2015. A 300-year surge his- tory of the Drangajökull ice cap, northwest Ice- land, and its maximum during the ”Little Ice Age”. Holocene 25(7), 1076–1092. http://doi.org/10.1177/- 0959683615576232

Harning, D. J., Á. Geirsdóttir, G. H. Miller and L. Anderson 2016. Episodic expansion of Drangajökull, Vestfirðir, Iceland, over the last 3 ka culminating in its maximum dimension during the Little Ice Age. Quat. Sci. Rev. 152, 118–131. http://doi.org/10.1016/- j.quascirev.2016.10.001

Extract from Brynjolfsson et al, 2016:

Surges of outlet glaciers from the Drangajökull ice cap, northwest Iceland,  Earth and Planetary Science LettersVolume 450, 15 September 2016, Pages 140-151

The reason for Drangajökull outlets displaying surge behaviour so different from surging glaciers of the other Icelandic ice caps remains to be fully understood. Bedrock substrate or flow rate has hitherto not been considered obvious variables controlling surge frequencies during the surge phase of Icelandic surge-type glaciers, although Brynjólfsson et al. (2015a) suggested it could not be excluded that the Miocene basalts below Drangajökull, which are different from the predominantly Pliocene–Pleistocene bedrock of other Icelandic ice caps, could at least partly explain the surge behaviour. Svalbard surging glaciers, recently suggested to be over 700 in number (Farnsworth et al., 2016), do occur on a variety of subglacial lithologies, from igneous and metamorphic Precambrian–Paleozoic basement rocks to sedimentary and igneous rocks of Paleozoic–Cenozoic age (Jiskoot et al., 1998, Murray et al., 2003, Farnsworth et al., 2016), but generally share the characteristics of long surge cycles and duration of the active surge phase. The geothermal heat flux below Drangajökull is comparable to the heat flux below Brúarjökull and Múlajökull (a very active surging outlet of Hofsjökull), estimated to be between 100–200 mW/m2 (Hjartarsson, 2015), so that the very different surge dynamics and kinematics can probably not be explained by differences in geothermal heat flux. Recent studies of Drangajökull have focused on geomorphology, glacial history since the Last Glacial Maximum (LGM), surge dynamics and recent areal changes of the ice cap (Principato, 2003, Principato, 2008, Principato et al., 2006, Þrastarson, 2006, Brynjólfsson et al., 2014, Brynjólfsson et al., 2015a, Brynjólfsson et al., 2015b). Differencing of Digital Elevation Models (DEMs) is a well-established methodology to quantify volume changes of glaciers (e.g., Magnússon et al., 2005, Magnússon et al., 2015, Schomacker and Kjær, 2007, Schomacker and Kjær, 2008, Sund et al., 2009, Sund et al., 2014, Abermann et al., 2010, Kjær et al., 2012, Schomacker et al., 2012, Jóhannesson et al., 2013). Time series of DEMs and other remotely sensed data are also commonly used to identify glacier surges and quantify their velocity, surface, volume and areal changes during the surges (Fischer et al., 2003, Magnússon et al., 2005, Magnússon et al., 2015, Frappé and Clarke, 2007, Sund et al., 2009, Sund et al., 2014, Quincey et al., 2011).
Shuman et al. (2009) compared a GPS derived Digital Elevation Model (DEM) with series of repeated satellite profiles across Drangajökull, indicating up to 1.5 m a−1 surface lowering at the location of the satellite profile in the years 2003–2007. However, ablation stake measurements indicate positive mass balance of the whole ice cap in 2005–2007, indicating that the satellite profile is not representative for the whole ice cap (Shuman et al., 2009).

Comparison of DEMs since c. 1990 and from 2011, indicates about 8 m average surface lowering of the ice cap in the period 1990–2011 (Jóhannesson et al., 2013). Obvious ice discharges during the most recent surges of the three outlet glaciers are reflected as much more thinning of their reservoirs and distinct thickening of the receiving areas of each outlet glacier as described by Jóhannesson et al. (2013). Recently, Magnússon et al. (2015) calculated the geodetic mass balance of Drangajökull in six time steps back to 1946 based on DEMs from aerial photographs. They demonstrated a mean mass balance rate of the ice cap of in the period 1946–2011. However, they also observed high decadal variability in the mass balance with, e.g., a positive mass balance rate of in 1975–1985 and from 1985–1994. Thus, the mass balance of Drangajökull does not appear to follow exactly the increasing temperature trend in the Arctic (Miller et al., 2010). However, on longer timescales, Drangajökull is thinning and its surface area has decreased since the LIA and during the last decades (Jóhannesson et al., 2013, Brynjólfsson et al., 2014, Brynjólfsson et al., 2015a, Magnússon et al., 2015).

The aim of this study is to quantify the ice elevation and volume changes related to the most recent surges (1994–2006) of the three major outlet glaciers of Drangajökull ice cap. Furthermore, we analyse the characteristics of Drangajökull surges and discuss their distinct difference from surges from the other Icelandic ice caps.

Setting
The Drangajökull ice cap is located c. 100–915 m above sea level (a.s.l.) on the eastern Vestfirðir (Westfjords) peninsula in northwest Iceland (Fig. 1). Since the LIA, the glaciated area has decreased from about 190–216 km2(Sigurðsson et al., 2013, Brynjólfsson et al., 2015a) to 142 km2 in 2011 (Jóhannesson et al., 2013). Drangajökull is located at the gateway to the Arctic, where the relatively warm Irminger branch of the North Atlantic Current and the cold East Greenland Current meet, and
Aerial photographs and Digital Elevation Models

Two sets of aerial stereo-photographs and derived DEMs from 1994 and 2005 in addition to an airborne Light Detection and Ranging (LiDAR) derived DEM from 2011, were used in this study (Table 1). The Root Mean Square Error (RMSE) of the DEMs (Table 1) is used to assign uncertainty in the elevation measurements. The 1994 aerial photographs were supplied by the National Land Survey of Iceland (Landmælingar Íslands). Orthorectified aerial photographs and a DEM with 5 m ground resolution were
Ice surface and volume changes 1994–2011

The three individual DEMs (Table 1) enable calculations of ice surface and volume changes for the period 1994–2011, 1994–2005 and 2005–2011 (Table 2). Because the DEM from 1994 does not completely cover Kaldalónsjökull, or its forefield (Fig. 4), Kaldalónsjökull was mostly excluded from this study. However, according to the 1994–2011 DoD (Fig. 4) it was only the western and northwest part of Kaldalónsjökull that contributed to its last surge in 1995–1999. The 1994–2011 DoD covers 102 km2 of the Reykjarfjarðarjökull.

The nunatak, Hljóðabunga (Fig. 1 and Fig. 2), seems to have acted as a barrier for ice flow during the surge of Reykjarfjarðarjökull 2002–2006. Longitudinal fractures and folded sediment bands within the ice are visible on aerial photo and can be traced from Hljóðabunga downglacier to the margin. Thus, the surge did not affect the glacier south of Hljóðabunga during the recent most surge in 2002–2006 (Fig. 1). GIS analysis indicates that most of the ice-volume discharge was from above the ELA

Conclusions
By using DEMs from 1994, 2005 and 2011 we have quantified ice surface elevation and volume changes that relate to the recent most surges of the surging outlet glaciers, Reykjarfjarðarjökull and Leirufjarðarjökull.

Holocene glacier and climate variations in Vestfirðir, Iceland, from the modeling of Drangajökull ice cap
2018, Quaternary Science Reviews

Citation Excerpt :

Because of the ice cap's narrow width, low surface slopes, and low maximum elevation, the rain shadow effect across Drangajökull is small (Roe and Baker, 2006), compared to the other large Icelandic ice caps (Einarsson, 1977; Crochet et al., 2007). Since 1930, the surge and outlet-length histories of Drangajökull were documented through direct observation and historical accounts (Sigurðsson, 1998; Björnsson et al., 2003; Brynjólfsson et al., 2014, 2015a; 2016; Ingólfsson et al., 2016; Magnússon et al., 2016a). Glaciers contracted since 1930, despite three outlets of Drangajökull each surging twice (Sigurðsson, 1998).

The Drangajökull ice cap, northwest Iceland, persisted into the early-mid Holocene
2016, Quaternary Science Reviews

Citation Excerpt :

Recently, Brynjólfsson et al. (2014, 2015a, b) outlined the history of Drangajökull outlets since the Little Ice Age as well as highlighting the surge-type outlet glacier dynamics. Brynjólfsson et al. (2014, 2016) and Ingólfsson et al. (2016) pointed out that the Drangajökull surge-type glaciers behaved more like the polythermal Svalbard surging outlets than the warm-based surging outlets of other Icelandic ice caps. Understanding the Holocene pattern of Drangajökull’s oscillations can improve our understanding of the dynamics of Holocene environmental changes in this key area.

Knob and kettle topography in Kaldalon valley

 

Here, in the valley of Kaldalon, in NW Iceland, which I have visited many times,  we have a text book example of knob and kettle topography. When we were working there, we referred to this as the Trout Pools area, since there were small trout in some of the little lakes in the depressions or "kettles". 

Some authors use the phrase "kame and kettle topography" -- presumably in fear of getting into trouble with the guardians of public morality.......... 

Compared with the abandoned outwash fan on its SW flank, the Trout Pools area is extremely complex, with sinious esker ridges, mounds of sand and gravel, dry channels cut by flowing meltwater, and abundant hollows -- some containing pools or lakes, and some dry.  Some researchers who have worked here refer to this area -- mistakenly -- as "pitted outwash".  Especially on the eastern part of the discussed terrain there is much till (in patches or hummocks) resting on top of the fluviogloacial materials, and there are some distinct mounds of reddish "rhyolite till".    It is clear therefore that at least some of the topographic features were formed sub-glacially -- and that some spreads of fluvioglacial sands and gravels were deposited on top of melting dead ice.

At the time of formation the ice edge must have been melting catastrophically from its maximum at the Trimbilsstadir  moraine -- widely assumed nowadays as having been formed at the peak of the 1740 glacier surge.

I shall do another post on the surges and glacier snout oscillations .........

PS (17 Dec 2025)

Following examination of more satellite images in different light conditions, and consulting my ancient field notes, I am coming round to the view that the "esker" shown on the inside of the green Trimbilsstadir Moraine is not as esker at all, but a ridge showing an ice-front position, with a very well defined ice-contact face on the up-valley side.  See my other posts on this.

Kaldalonsjökull -- emergence of a hidden landscape

 

Satellite image of the 2025 Kaldalonsjökull.  This is a composite image with some rough joins.......

Two annotated close-ups from the same satellite imagery.

Sad though it is to comment on the death of a glacier, it is fascinating to record the emergence of a landscape that might not have seen daylight for more than a hundred thousand years. On the other hand, there is an active debate in the literature about the survival or disappearance of Drangajokull and its outlet glaciers duringt the Holocene.  The drift of specialist opinion now seems to be that the ice cap melted away during the Holocene and was absent duting the "Neoglacial" period, to be regenerated around 2,000 years ago, reaching its greatest extent in the Litrtle Ice Age.  Watch this space........

 This is a high resolution satellite image from Google Maps, taken this year.  If you zoom in even closer, you can see the individual boulders littering the ground surface.........

As I pointed out in a previous post, this is a complex landscape of rocky knolls, platforms controlled by flat-lying basalts, wide gorges and narrow meltwater channels.  Some of the undulating terrain still supports patches of dead ice -- elsewhere the ice has gone altogether, to be replaced nowadays just by a seasonal snowcover which changes from year to year in respose to precipitation totals and the directions of snowdrifting.  There are traces of pitted moraine, fluted moraine and ridged moraine, probably related to recent glaciological conditions -- ie ice wastage within the last few decades.

One of our photos from 1960.  The beginning of the end.......

Dave at the exit of the Morilla River from the glacier snout.

If Kaldalonsjökull really is only 2,000 years old, then we can infer that the "hidden landscape" represents a complex history of  waxing and waning, disappearance and regeneration.  In other words, there have been multiple phases of growth and decline, including catastrophic ice wastage and powerful meltwater flow at times.  No wonder this hidden landscape is so complex....... 

===============

Abstract
The status of Icelandic ice caps during the early Holocene provides important constraints on North Atlantic climate and the mechanisms behind natural climate variability. A recent study postulates that Drangajökull on Vestfirðir, Iceland, persisted through the Holocene Thermal Maximum (HTM, 7.9–5.5 ka) and may be a relic from the last glacial period. We test this hypothesis with a suite of sediment cores from threshold lakes both proximal and distal to the ice cap's modern margin. Distal lakes document rapid early Holocene deglaciation from the coast and across the highlands south of the glacier. Sediment from Skorarvatn, a lake to the north of Drangajökull, shows that the northern margin of the ice cap reached a size comparable to its contemporary limit by ∼10.3 ka. Two southeastern lakes with catchments extending well beneath modern Drangajökull confirm that by ∼9.2 ka, the ice cap was reduced to ∼20% of its current area. A continuous 10.3ka record of biological productivity from Skorarvatn's sediment indicates local peak warmth occurred between 9 and 6.9 ka. The combination of warm and dry summers on Vestfirðir suggests that Drangajökull very likely melted completely shortly after 9.2 ka, similar to most other Icelandic ice caps.

from Harning et al, 2016:

https://www.sciencedirect.com/science/article/abs/pii/S0277379116303924?via%3Dihub

Early Holocene deglaciation of Drangajökull, Vestfirðir, Iceland
October 2016
Quaternary Science Reviews 153
DOI:
10.1016/j.quascirev.2016.09.030

Kaldalonsjökull -- portrait of a dying glacier

When we worked on this glacier in 1960 the snout was located in the foreground of the photo.  In 65 years there has been a phenomenal retreat of over 1 km, exposing the rocky knolls to the right and the spectacular rock bench or ledge that we see in the centre of the image.  However, this retreat from the 1940 surge position was disrupted by the most recent surge between 1995 and 1999, which involved a readvance of 1015m.  Over the past 25 years there has been a continuous retreat of the ice edge.

I have just come across this fascinating photo collection, with recent photosn of the snout of the Kaldalonsjökull -- one of the small outlet glaciers associated with the Drangajokull ice cap in NW Iceland.

https://icelandthebeautiful.com/kaldalonsjokull-drangajokull-strandir-vestfirdir-island/#:~:text=Kaldal%C3%B3nsj%C3%B6kull%20is%20a%20glacier%20tongue,falling%20rocks%2C%20so%20be%20cautious.

With colleagues, I worked on this glacier in 1960 and 1973-74, when it was quite active, covering virtually all of the rock exposures seen in the above photo:

Kaldalonsjokull snout, 1960.

The current snout, which is nothing more than a patch of dead ice detatched from the main ice cap.  The trough head cliff at the top of the photo is a prominent feature.

Another view of the present-day wasting glacier, with just one current connection to the ice cap, on the extreme right of the image.

The evolving situation

The 2006 satellite image.  The position of the rock ledge is shown by the snow banked up against the cliffline.  In the 1970's the rock ledge was occasionally exposed but mostly snow-covered.

Left image: 1960 aerial photo. Right image: 2025 satellite image. there has been an overall retreat of over 1 km.  However, in the period between 1960 and 2025 there has been another surge which pushed the ice front back to approximately its 1960 position, leading to a somewhat confusing situation in the current ice wastage zone........

View of the upper part of the glacier, seen from the adjacent Votubjorg basalt plateau, 1975.  This is the position of the trough head or bench, with a readily identifiable cliff edge and an extensive platform with fluted till and a scatter of boulders above it.  Since 1960 this cliff edge has sometimes been visible and sometimes not -- depending on the cycle of surges and ice wastage episodes.

The pattern of surges and retreats on Kaldalonsjökull from 1740 to the present day.  After Brynjolfsson et al, 2015.  Note the most recent surge of 1 km betwewen 1995 and 1999, which has been followed by a very rapid ice edge retreat.

Between 1940 and 1994 there was a continuous retreat of the ice edge of 1.5 km;  when we first visited in 1960 the effects of this ongoing retreat were clearly displayed.  The surge of 1994-99 pushed the ice edge forward by 1015m, and the rapid retreat that followed has seen the ice edge retreat further into the trough head than the 1994 position.

Brynjólfsson, S., A. Schomacker, E. R. Guðmundsdóttir and Ó. Ingólfsson. 2015. A 300-year surge his- tory of the Drangajökull ice cap, northwest Ice- land, and its maximum during the ”Little Ice Age”. Holocene 25(7), 1076–1092. http://doi.org/10.1177/- 0959683615576232

Abstract

Over the last 300 years, each of the three surge-type outlet glaciers of the Drangaj.kull ice cap in northwest Iceland has surged 2–4 times. There is valuable historical information available on the surge frequencies since the ‘Little Ice Age’ (LIA) maximum because of the proximity of the surging outlets,
Reykjarfjar.arj.kull, Leirufjar.arj.kull and Kaldal.nsj.kull, to farms and pastures and monitoring of these glaciers since 1931. We have reconstructed the surge history of the Drangaj.kull ice cap, based on geomorphological mapping, sedimentological studies and review of historical records. Geomorphological mapping of the glacier forefields reveals twice as many end moraines as previously recognized. This indicates a higher surge interval than earlier perceived. A clear relationship between the surge interval and climate cannot be established. Surges were observed more frequently during the 19th century and the earliest 20th century compared with the relatively cool 18th century and the late 20th century, possibly reflecting a lack of information rather than a long quiescent phase of the glaciers. We have estimated the magnitude of the maximum surge events during the LIA by reconstruction of Digital Elevation Models (DEMs) that can be compared with modern DEMs. As reference points for the digital elevation modelling, we used the recently mapped lateral moraines and historical information on the exposure timing of nunataks. During the LIA maximum surge events, the outlet glaciers extended 3–4 km further down-valley than at present. Their ice volumes were at least 2–2.5 km3 greater than in the beginning of the 21st century.