A multi-proxy approach to Late Holocene fluctuations of 1 

Tungnahryggsjökull glaciers in the Tröllaskagi peninsula (northern 2 

Iceland) 3 
 4 
José M. Fernández-Fernández a *, David Palacios a, Nuria Andrés a, Irene Schimmelpfennig b, 5 
Skafti Brynjólfsson c, Leopoldo G. Sancho d, José J. Zamorano e, Starri Heiðmarsson c, Þorsteinn 6 
Sæmundsson f, ASTER Team b 1 7 
 8 
a High Mountain Physical Geography Research Group, Department of Geography, Faculty of Geography 9 
and History, Universidad Complutense de Madrid, 28040 Madrid, Spain 10 
b Aix-Marseille Université, CNRS, IRD, Coll. France, UM 34 CEREGE, Technopôle de l'Environnement 11 
Arbois-Méditerranée, BP 80, 13545 Aix-en-Provence, France 12 
c Icelandic Institute of Natural History, Borgum vid Norðurslóð, Box 180, 602 Akureyri, Iceland 13 
d Departament of Vegetal Biology II, Faculty of Pharmacology, Universidad Complutense de Madrid, 28040 14 
Madrid, Spain 15 
e Instituto de Geografía, Universidad Nacional Autónoma de México, Ciudad Universitaria, 04510 Ciudad 16 
de México, Mexico 17 
f Faculty of Life and Environmental Science, University of Iceland, Öskju, Sturlugötu 7, 101 Reykjavík, 18 
Iceland 19 
* Corresponding author: José M. Fernández-Fernández (josemariafernandez@ucm.es) 20 

 21 

Abstract 22 

The Tröllaskagi Peninsula in northern Iceland hosts more than a hundred small glaciers that have 23 
left a rich terrestrial record of Holocene climatic fluctuations in their forelands. Traditionally, it 24 
has been assumed that most of the Tröllaskagi glaciers reached their Late Holocene maximum 25 
extent during the Little Ice Age (LIA). However, there is evidence of slightly more advanced pre-26 
LIA positions. LIA moraines from Iceland have been primary dated mostly through lichenometric 27 
dating, but the limitations of this technique do not allow dating of glacial advances prior to the 28 
18th or 19th centuries. The application of 36Cl Cosmic-Ray Exposure (CRE) dating to 29 
Tungnahryggsjökull moraine sequences in Vesturdalur and Austurdalur (central Tröllaskagi) has 30 
revealed a number of pre-LIA glacial advances at ~400 and ~700 CE, and a number of LIA 31 
advances in the 15th and 17th centuries, the earliest LIA advances dated so far in Tröllaskagi. This 32 
technique hence shows that the LIA chronology in Tröllaskagi agrees with that of other European 33 
areas such as the Alps or the Mediterranean mountains. The combined use of lichenometric dating, 34 
aerial photographs, satellite images and fieldwork shows that the regional colonization lag of the 35 
commonly used lichen species Rhizocarpon geographicum is longer than previously assumed. 36 
For exploratory purposes, an alternative lichen species (Porpidia soredizodes) has been tested for 37 
lichenometric dating, estimating a tentative growth rate of 0.737 mm yr−1. 38 

 39 

Keywords 40 

Iceland; Tröllaskagi; Late Holocene glacier evolution; Little Ice Age; 36Cl Cosmic-Ray Exposure 41 
dating; Lichenometric dating 42 

mailto:josemariafernandez@ucm.es


1. Introduction 43 

The Tröllaskagi Peninsula (northern Iceland) hosts over 160 small alpine cirque glaciers 44 
(Björnsson, 1978; see synthesis in Andrés et al., 2016). Only a few of these small glaciers do not 45 
have supraglacial debris cover that allows them to react quickly to small climatic fluctuations 46 
(Caseldine, 1985b; Häberle, 1991; Kugelmann, 1991), compared to the reduced dynamism of the 47 
predominant rock glaciers and debris-covered glaciers (Martin et al., 1991; Andrés et al., 2016; 48 
Tanarro et al., 2019). As a result of their high sensitivity to climatic changes, the few debris-free 49 
glaciers in Tröllaskagi fluctuated repeatedly in the past, forming a large number of moraines in 50 
front of their termini (see Caseldine, 1983, Caseldine, 1985b, Caseldine, 1987; Kugelmann, 1991 51 
amongst others). However, the relation between glacier fluctuations and the climate is 52 
complicated due to: (i) the well-known surging potential activity in Tröllaskagi (Brynjólfsson et 53 
al., 2012; Ingólfsson et al., 2016); (ii) the possibility of glaciers being debris-covered in the past, 54 
with the subsequent change of their climate sensitivity over time; (iii) the intense and dynamic 55 
slope geomorphological activity, paraglacial to a great extent (Jónsson, 1976; Whalley et al., 56 
1983; Mercier et al., 2013; Cossart et al., 2014; Feuillet et al., 2014; Decaulne and Sæmundsson, 57 
2006; Sæmundsson et al., 2018) that hides and erases the previous glacial features. 58 

In Iceland, a great part of the research on glacial fluctuations during the late Holocene, LIA 59 
(Grove, 1988; 1250–1850 CE, see Solomina et al., 2016) and later stages has been approached 60 
through application of radiocarbon dating of organic material, analyzing lake sediment varves 61 
(Larsen et al., 2011; Striberger et al., 2012), dead vegetation remnants (Harning et al., 2018) and 62 
threshold lake sediment records (Harning et al., 2016b; Schomacker et al., 2016). These provide 63 
high-resolution records of glacier variability. In Tröllaskagi, very few reliable dates exist at 64 
present, obtained from radiocarbon and tephrochronology (Häberle, 1991, Häberle, 1994; Stötter, 65 
1991; Stötter et al., 1999; Wastl and Stötter, 2005). They suggest that late Holocene glaciers were 66 
slightly more advanced than during the LIA maximum in Tröllaskagi. However, these techniques 67 
only provide minimum or maximum ages for Holocene glacier history in northern Iceland (Wastl 68 
and Stötter, 2005). In addition, at Tröllaskagi most of the moraines are at 600–1000 m a.s.l., where 69 
the applicability of tephrochronology is very limited (Caseldine, 1990) due to the great intensity 70 
of slope geomorphological processes. 71 

In any case, most of the moraine datings of Tröllaskagi –especially those of the last millenium– 72 
comes from lichenometry (see synthesis in Decaulne, 2016). However, the ages derived from this 73 
technique tend to be younger if they are compared to those estimated from tephra layers 74 
(Kirkbride and Dugmore, 2001). Generally, there is disagreement about the validity of the results 75 
provided by lichenometry, due to the difficulty of the lichen species identification in the field, the 76 
complexity and reliability of the sampling and measurement strategies, the uncertainty estimates, 77 
the nature of the ages provided as relative, as well as the relative reliability at producing lichen 78 
growth curves of the different species, with an extreme dependence on local environmental factors 79 
(see Osborn et al., 2015). 80 

In fact, these problems had already been detected in northern Iceland, where in spite of the time 81 
elapsed since the first applications of this technique (Jaksch, 1970, Jaksch, 1975, Jaksch, 1984; 82 
Gordon and Sharp, 1983; Maizels and Dugmore, 1985) and contrasting with its widespread use 83 
in the south and southeast of the island (Thompson and Jones, 1986; Evans et al., 1999; Russell 84 
et al., 2001; Kirkbride and Dugmore, 2001; Bradwell, 2001, Bradwell, 2004a; Harris et al., 2004; 85 
Bradwell and Armstrong, 2006; Orwin et al., 2008; Chenet et al., 2010 and others), there are still 86 
very few established growth rates for the lichen group Rhizocarpon geographicum (Roca-Valiente 87 



et al., 2016) from Tröllaskagi (Caseldine, 1983; Häberle, 1991; Kugelmann, 1991; Caseldine and 88 
Stotter, 1993; see synthesis in Decaulne, 2016). In general, the growth curves from Tröllaskagi 89 
suffer from a few control points (i.e. surfaces of known age) for calibration. This leads to 90 
considerable underestimation when lichen thalli sizes are beyond the calibration growth curve 91 
(Caseldine, 1990; see e.g. Caseldine, 1985b; Caseldine and Stotter, 1993; Kugelmann, 1991). That 92 
is to say, the extrapolation does not account for the decreasing growth rate with increasing age 93 
(i.e. non-linear growth). Kugelmann's (1991) growth curve has the highest number of control 94 
points to date (19 in total). In addition, the colonization lag in Tröllaskagi is poorly defined, 95 
although 10–15 years have been assumed for the Rhizocarpon geographicum (Caseldine, 1983; 96 
Kugelmann, 1991). Other issues, such as the absence of large thalli (Maizels and Dugmore, 1985), 97 
lichen saturation (e.g. Wiles et al., 2010), and other environmental factors (Innes, 1985 and 98 
references included; Hamilton and Whalley, 1995) contribute to appreciable age underestimations 99 
when dating surfaces using this approach. Snow is also a major environmental factor for 100 
lichenometry in Iceland due to its long residence time on the ground (Dietz et al., 2012) and the 101 
avalanching frequency –especially in Tröllaskagi–, whose effect restricts the growth rate of 102 
lichens, and even destroys them (Sancho et al., 2017). These problems considerably limit the 103 
quality (reliability) of the lichenometric dating range of utility for applying lichenometric dating 104 
in Iceland, where the oldest ages estimated so far are between 160 and 220 years (Maizels and 105 
Dugmore, 1985; Thompson and Jones, 1986; Evans et al., 1999), thus preventing the dating of 106 
glacial advances prior to the 18th century. 107 

In spite of the high uncertainty of the lichenometric dating, many authors working in Tröllaskagi 108 
have treated their lichenometric results as absolute ages (see Caseldine, 1983, Caseldine, 1985b, 109 
Caseldine, 1987; Kugelmann, 1991; Häberle, 1991; Caseldine and Stotter, 1993, amongst others). 110 
In fact, previously considered dates of LIA maximum glacier culmination are restricted to the 111 
very late 18th and early 19th centuries (Caseldine, 1983, Caseldine, 1985b, Caseldine, 1987; 112 
Kugelmann, 1991; Caseldine and Stotter, 1993; Caseldine, 1991; Martin et al., 1991), very close 113 
to the applicability threshold of this method. Consequently, no evidence of previous advances 114 
during phases of the LIA that were more conducive to glacier expansion (probably colder; see e.g. 115 
Ogilvie, 1984, Ogilvie, 1996; Ogilvie and Jónsdóttir, 2000; Ogilvie and Jónsson, 2001) has been 116 
found (Kirkbride and Dugmore, 2001). 117 

In the recent years, dating methods based on the Exposure to the Cosmic-Rays (CRE) have been 118 
introduced successfully to date moraines of the last millennium, and even formed during the LIA 119 
(Schimmelpfennig et al., 2012, Schimmelpfennig et al., 2014a, Schimmelpfennig et al., 2014b; 120 
Le Roy et al., 2017; Young et al., 2015; Jomelli et al., 2016; Li et al., 2016; Dong et al., 2017; 121 
Palacios et al., 2019). The cosmogenic nuclides 36Cl and 3He have been applied previously in 122 
northern Iceland to date the Pleistocene deglaciation. Principato et al. (2006) studied the 123 
deglaciation of Vestfirðir combining 36Cl CRE dating of moraine boulders and bedrock surfaces 124 
with marine records and tephra marker beds. Andrés et al. (2019) reconstructed the deglaciation 125 
at the Late Pleistocene to Holocene transition at Skagafjörður through 36Cl CRE dating applied to 126 
polished surfaces along a transect from the highlands to the mouth of the fjord. Brynjólfsson et 127 
al. (2015b) applied the same isotope to samples coming from the highlands and the fjords to 128 
reconstruct the glacial history of the Drangajökull region. The other cosmogenic isotope used is 129 
3He, applied to helium-retentive olivine phenocrysts by Licciardi et al. (2007) to determine 130 
eruption ages of Icelandic table mountains and to reconstruct the volcanic history and the 131 
thickness evolution of the Icelandic Ice Sheet during the last glacial cycle. However, CRE dating 132 
has not yet been applied to the late Holocene glacial landforms both in Iceland as a whole and the 133 



Tröllaskagi Peninsula. CRE dating is an alternative to the use of high-resolution continuous 134 
lacustrine records in northen Iceland, given the rarity of lakes in this peninsula, which limits the 135 
application of radiocarbon to date the deglaciation processes (Striberger et al., 2012; Harning et 136 
al., 2016b). 137 

Nevertheless, CRE dating methods approach the nuclides' detection limit when applied to very 138 
recent moraines (Marrero et al., 2016; Jomelli et al., 2016). This issue precludes the application 139 
of CRE dating to the abundant post-LIA moraines existing in some of the Tröllaskagi valleys 140 
whose headwalls are occupied by climate-sensitive debris-free glaciers (Caseldine and 141 
Cullingford, 1981; Caseldine, 1983, Caseldine, 1985b; Kugelmann, 1991; Fernández-Fernández 142 
et al., 2017). Dating these post-LIA moraines allows to reconstruct recent climate evolution, and 143 
even to match and assess the climate reconstructions with the instrumental climate records (see 144 
Dahl and Nesje, 1992; Caseldine and Stotter, 1993; Fernández-Fernández et al., 2017). 145 
Furthermore, improving the knowledge of the recent evolution of alpine mountain glaciers, such 146 
as those of Tröllaskagi, is fundamental in the assessment of present global warming (Marzeion et 147 
al., 2014). 148 

The use of aerial photographs from different dates is a reliable approach to the glacier evolution 149 
during the last decades (Fernández-Fernández et al., 2017; Tanarro et al., 2019). The main 150 
advantage of this technique is the possibility of studying the evolution of glacier snouts in recent 151 
dates with high accuracy. In fact, there is no dependence on the glacial features (i.e. moraines), 152 
which circumvents moraine deterioration issues derived by the geomorphological activity of the 153 
slopes. However, the main shortcoming of the aerial photo imagery is the availability only on few 154 
dates, at least in Tröllaskagi. This circumstance makes it impossible to obtain the glacier 155 
fluctuations with a high time resolution (i.e. only the periods between the available aerial photos; 156 
Fernández-Fernández et al., 2017; Tanarro et al., 2019), and hence to match them to short-term 157 
(decadal scale) climate fluctuations that are known to exert a major control especially on small 158 
mountain glaciers with short time responses (see Caseldine, 1985b; Sigurðsson, 1998; Sigurðsson 159 
et al., 2007; Fernández-Fernández et al., 2017). The only way to fill the gap between two dates 160 
with available aerial photos is through applying lichenometric dating (Sancho et al., 2011), as 161 
there is no tree species suitable to apply dendrochronology. In addition, the information provided 162 
by the aerial photo imagery (i.e. glacier snout position) constrains the period when the lichens 163 
appear and begin to grow, which circumvents many of the criticisms made on lichenometry (see 164 
Osborn et al., 2015). 165 

Moreover, a detailed geomorphological mapping allows for identification of stable moraines, not 166 
remobilized or destroyed by glacier advances or slope processes (avalanches, slope deformations, 167 
debris-flows, etc.) and also to reconstruct the glacier snout geometry throughout different phases 168 
(see Caseldine and Cullingford, 1981; Bradwell, 2004b; Principato et al., 2006 amongst others). 169 
The analysis of the moraine morphology and the glacial features on the forelands is a key tool to 170 
confirm whether the glaciers were debris-free or debris-covered in the past (Kirkbride, 2011; 171 
Janke et al., 2015; Knight et al., 2018, amongst others) as this issue determines their climate 172 
sensitivity (see Fernández-Fernández et al., 2017; Tanarro et al., 2019). 173 

The western and eastern Tungnahryggsjökull glaciers, in the Vesturdalur and Austurdalur Valleys 174 
(central Tröllaskagi), respectively (Fig. 1), are two of the few debris-free glaciers –or almost 175 
debris-free in the case of the western glacier– of the peninsula that are both small and highly 176 
sensitive to climatic fluctuations (Fernández-Fernández et al., 2017). This makes them ideal for 177 
studying glacial and climatic evolution during last millennia. 178 



The aim of our work was to apply the best methodology possible to analyze the glacial evolution 179 
of western and eastern Tungnahryggsjökull glaciers during the last millennia to the present. 180 
Applying for the first time a number of dating techniques to study the Late Holocene evolution 181 
of the two glaciers, the objectives of this paper are: 182 

(i) To carry out a detailed geomorphological survey of the glacier forelands in order to 183 
map accurately well-preserved glacial features. This mapping is used both to devise 184 
the sampling strategy for dating, and also to reconstruct the palaeoglaciers in 3D in 185 
order to obtain glaciologic climate indicators such as the Equilibrium-Line Altitudes 186 
(ELAs). These can be used as a proxy to infer palaeoclimatic information (see Dahl 187 
and Nesje, 1992; Caseldine and Stotter, 1993; Brugger, 2006; Hughes et al., 2010; 188 
Fernández-Fernández et al., 2017 amongst others). 189 

(ii) To use aerial photographs/satellite imagery post-dating 1946 to map the glacier extent 190 
in each available date in order to improve the information of ELA evolution in the 191 
recent decades. Aerial photographs will be used to constrain the possible periods of 192 
lichen colonization and growth over stable boulders, and will be useful to identify 193 
phases of advance, stagnation or retreat of the glacier snouts. By this way we will 194 
complete the glacier evolution from pre-instrumental glacial stages (identified from 195 
geomorphological evidence, i.e. moraines) to their current situation. 196 

(iii) To apply CRE dating when possible depending on the preservation degree of the 197 
glacial features, and when moraines were too old to be dated by lichenometry. 198 

(iv) To apply lichenometric relative dating to recent moraines or those where CRE dating 199 
might not be suitable (i.e. limit of applicability) and provided that: 1) the 200 
geomorphological criteria evidence a good preservation of glacial features; and 2) 201 
aerial photographs constrain the earliest and oldest possible lichen ages. This 202 
approach will also allow checking the growth rates and colonization lags of the lichen 203 
species usually used in Iceland for lichenometric dating purposes. 204 

The experimentation and validation of this methodological purpose will help to improve the 205 
knowledge of the recent climate evolution of northern Iceland. This is of maximum interest if we 206 
consider its location within the current atmospheric and oceanic setting, strongly linked to the 207 
evolution of the Meridional Overturning Circulation (Andrews and Giraudeau, 2003; Xiao et al., 208 
2017), a key factor in the studies for the assessment of the global climate change effects (Barker 209 
et al., 2010; Chen et al., 2015 amongst others). Moreover, if this proposal is valid, it could be 210 
applied to the research on the recent evolution of other mountain glaciers similar to those of 211 
Tröllaskagi. This aspect is a main research objective at present, as these glaciers represent the 212 
greatest contribution to the current sea-level rise (Jacob et al., 2012; Gardner et al., 2013 amongst 213 
others). 214 

2. Regional setting 215 

The Tröllaskagi Peninsula extends into the Atlantic Ocean at 66°12′ N from the central highland 216 
plateau (65°23′ N) of Iceland (Fig. 1). The fjords of Skagafjörður (19°30′ W) and Eyjafjörður 217 
(18°10′ W) separate it from the Skagi and Flateyjarskagi peninsulas, respectively. Tröllaskagi is 218 
an accumulation of successive Miocene basalt flows, interspersed with reddish sedimentary strata 219 
(Sæmundsson et al., 1980; Jóhannesson and Sæmundsson, 1989). The plateau culminates at 220 
altitudes of 1000–1500 m a.s.l. (with the highest peak Kerling at 1536 m a.s.l.) and is dissected by 221 



deep valleys with steep slopes whose headwaters are currently glacial cirques. These cirques host 222 
>160 small glaciers, mostly north-facing, resulting from the leeward accumulation of snow from 223 
the plateau (Caseldine and Stotter, 1993) and reduced exposure to solar radiation. In fact, deposits 224 
caused by rock-slope failure are common in Tröllaskagi valley slopes and have been considered 225 
a result of the final deglaciation during the early Holocene (Jónsson, 1976; Feuillet et al., 2014; 226 
Cossart et al., 2014; Coquin et al., 2015). Most of the glaciers are rock or debris-covered glaciers, 227 
due to the intense paraglacial activity affecting the walls that minimizes cosmogenic nuclide 228 
concentrations from earlier exposure periods on the cirque headwalls (Andrés et al., 2019). 229 

The climate in Tröllaskagi is characterized by a mean annual air temperature (MAAT; 1901–1990 230 
series) of 2 to 4 °C at sea level, dropping to between −2 and − 4 °C at the summits (Etzelmüller 231 
et al., 2007). At the town of Akureyri, located in the east of the peninsula at the mouth of 232 
Eyjafjarðardalur (Fig. 1), the MAAT is 3.4 °C and the average temperature in the three summer 233 
months is about 9 to 10 °C (Einarsson, 1991). Annual precipitation (1971–2000) ranges from 234 
400 mm at lower altitudes to 2500 mm at the summits (Crochet et al., 2007). 235 

The frontier location of Icelandic glaciers in relation to atmosphere and ocean systems 236 
(warm/moist Subtropical and cold/dry Arctic air masses; and the warm Irminger and cold East 237 
Greenland sea currents) makes them exceptionally sensitive to climate oscillations (Bergþórsson, 238 
1969; Flowers et al., 2008; Geirsdóttir et al., 2009), and these debris-free glaciers are thus reliable 239 
indicators of climatic evolution and the impact of climate change on the cryosphere (see 240 
Jóhannesson and Sigurðsson, 1998; Bradwell, 2004b; Sigurðsson, 2005; Geirsdóttir et al., 2009; 241 
Fernández-Fernández et al., 2017). 242 

The glaciers studied here are the western (6.5 km2) and eastern (3.9 km2) Tungnahryggsjökull, 243 
located in the Vesturdalur and Austurdalur valleys respectively, separated from each other by the 244 
crest of Tungnaryggur, and tributaries of the Kolbeinsdalur Valley (Fig. 1). 245 

3. Methodology 246 

3.1. Geomorphological mapping and glacier geometry mapping 247 

Four summer field campaigns (2012, 2013, 2014 and 2015) were conducted in Vesturdalur and 248 
Austurdalur, with the objective of identifying moraines that clearly provided evidence of various 249 
glacial culminations of the western and eastern Tungnahryggsjökull glaciers. We identified the 250 
palaeo-positions of the glacier snouts, the glacier geometry and extent through photo 251 
interpretation of stereoscopic pairs and previous fieldwork. Mapping of glacial and non-glacial 252 
landforms was conducted on two enlarged 50-cm-resolution aerial orthophotos (National Land 253 
Survey of Iceland, 2018) plotted at scale ≈1:7000. These maps were imported into an ArcGIS 254 
1.4.1 database after geo-referencing. Finally, all the glacial linear landforms were digitized, and 255 
where the moraines were prominent, continuous, or well-preserved, they were assumed to 256 
represent major culminations, and hence, glacial stages. In addition to the geomorphological 257 
evidence, glacier variations in recent years were mapped, based on the photo interpretation of four 258 
historical aerial photographs from 1946, 1985, 1994 and 2000 (National Land Survey of Iceland, 259 
2018), and also one SPOT satellite image (2005), previously geo-referenced in ArcGIS. For more 260 
details of the aerial photograph processing, see Fernández-Fernández and Andrés (2018). As the 261 
glacier headwalls are ice diffluences in both cases, it was assumed that: (i) the ice divides are the 262 
upper boundaries of the glaciers for the different stages, and (ii) the ice divide is invariant for the 263 
different stages/dates, following Koblet et al. (2010) as changes in the extent of the accumulation 264 
zone are smaller than the outlining differences derived from the operator mapping. Likewise, in 265 



the case of the upper glacier edge constrained by the cirque wall, the glacier geometry was also 266 
assumed to be invariant from stage to stage unless the aerial photographs showed otherwise. 267 

3.2. Glacier reconstruction and Equilibrium-Line Altitude (ELA) calculation 268 

Benn and Hulton's (2010) glacier reconstruction approach using a physical-based model 269 
describing ice rheology and glacier flow (Van der Veen, 1999) was preferred to arbitrary hand-270 
drawn contouring (e.g. Sissons, 1974) in order to achieve more robust reconstructions 271 
(Fernández-Fernández and Andrés, 2018). This model operates on deglaciated areas with non-272 
extant glaciers. As this is not the case in our study area, we approached the glacier bedrock by 273 
tentatively estimating the ice thickness' spatial distribution on the two glaciers studied. The 274 
“VOLTA” (“Volume and Topography Automation”) ArcGIS toolbox (James and Carrivick, 2016) 275 
was applied with the default parameters. It only requires the glacier outline and its digital 276 
elevation model (DEM). In the first step, the tool “volta_1_2_centreline” creates the glacier 277 
centrelines, and then the tool “volta_1_2_thickness” estimates ice thickness at points along them, 278 
assuming perfect-plasticity rheology, and interpolates the values using a glaciologically correct 279 
routine. The final result is an ice-free DEM in which we reconstructed the glacier. The former 280 
glacier surface topographies at the different stages/dates were reconstructed applying the semi-281 
automatic “GLaRe” ArcGIS toolbox designed by Pellitero et al. (2016), which implements the 282 
Benn and Hulton (2010) numerical model and estimates ice-thickness along glacier flowlines. To 283 
simplify the glacier surface modelling, shear-stress was assumed to be constant along the glacier 284 
flowline and over time. Using the value 110 kPa for the shear-stress resembles best the current 285 
longitudinal profile of the glaciers in the Benn and Hulton (2010) spreadsheet. This value can be 286 
considered appropriate as it falls within the normal shear-stress range of 50–150 kPa observed in 287 
current glaciers and is very close to the standard value of 100 kPa (Paterson, 1994). The glacier 288 
contours were manually adjusted to the ice surface elevation values of the ice-thickness points 289 
estimated by “GLaRe” instead of using an interpolation routine, to obtain a more realistic surface 290 
(concave and convex contours above and below the ELA). 291 

Finally, we calculated the ELAs automatically by using the “ELA calculation” ArcGIS toolbox 292 
(Pellitero et al., 2015). The methods comprised: (i) AABR (Area Altitude Balance Ratio; 293 
Osmaston, 2005) with the ratio 1.5 ± 0.4 proposed for Norwegian glaciers by Rea (2009) and 294 
successfully tested on Tröllaskagi debris-free glaciers in Fernández-Fernández et al. (2017); and 295 
(ii) the AAR (Accumulation Area Ratio) with the ratio 0.67 previously used by Stötter (1990) and 296 
Caseldine and Stotter (1993) for Tröllaskagi glaciers. Alternative approaches for ELA calculation 297 
in northern Iceland have been carried out by considering morphometric parameters of glacial 298 
cirques (altitude ratio, cirque floor, minimum point; Ipsen et al., 2018). However, we preferred 299 
the methods considering the glacier hypsometry as they reflect more evident changes from stage 300 
to stage (Fernández-Fernández et al., 2017; Fernández-Fernández and Andrés, 2018), if compared 301 
to morphometric parameters of the cirques. In this sense, the cirque floor elevation is derived from 302 
the last erosion period, and it is impossible to know the values corresponding to previous glacial 303 
stages. 304 

3.3. Lichenometric dating procedures 305 

Lichenometry was used as a relative dating tool, assuming that the lichens increase in diameter 306 
with respect to age. The results aim to complete the age control (of recent landforms non suitable 307 
for CRE dating) on the periods between aerial photos of known date as it has been applied 308 
successfully to control lichen (Sancho et al., 2011) and bryophyte (Arróniz-Crespo et al., 2014) 309 
growth during primary succession in recently deglaciated surfaces. First, we surveyed the 310 



moraines thoroughly, starting from the current glacier snouts downwards, looking for large stable 311 
boulders (i.e. well embedded in the moraine, not likely of having been overturned or remobilized 312 
by slope processes which could have affected lichen growth, e.g. snow avalanches, debris-flows, 313 
rockfall, landslides or debris-flows) with surfaces valid for dating (not weathered or resulting 314 
from block break). Lichenometry was applied to date moraine ridge boulders with the following 315 
criteria and assumptions: (i) boulders must clearly belong to the moraine ridge; (ii) lichen species 316 
should be abundant enough to allow measurements of a number of thalli at each location and 317 
hence enable surfaces to be dated under favourable environmental conditions such as basaltic 318 
rocks in subpolar mountains; (iii) only the largest lichen (circular or ellipsoidal thalli) of species 319 
Rhizocarpon geographicum, located on smooth horizontal boulder surfaces, was measured; (iv) 320 
the lichenometric procedures should not be applied when the lichen thalli coalesce on the boulder 321 
surface and individual thalli cannot be identified. We preferred the geomorphological criterion 322 
(stability vs. slope processes) to the establishment of lichenometry plots of a fixed area (e.g. 323 
Bickerton and Matthews, 1992) to ensure that lichens were measured on reliable boulders. This 324 
measurement strategy tries to circumvent or at least to minimize the specific problems of 325 
Tröllaskagi when dating glacial features so that: (i) snow accumulation should be lower in the 326 
moraine crest; (ii) lichen ages will be estimated only for the boulders located on the crests used 327 
to map and reconstruct the glaciers; and (iii) and lichens subjected to thalli saturation 328 
(coalescence) or high competition are not measured. 329 

First, Rhizocarpon geographicum lichens were measured with a Bernier calibrator. Then, digital 330 
photographs for high-precision measurements were taken of the most representative of the largest 331 
thallus located in each selected boulder (Suppl. Fig. SF1), using an Icelandic króna coin (21 mm 332 
diameter) parallel to the surface of the lichen as a graphic scale. We preferred the single largest 333 
lichen approach as previously has been done in Tröllaskagi for lichenometric dating of moraines 334 
(Caseldine, 1983, Caseldine, 1985b, Caseldine, 1987; Kugelmann, 1991). The photos were scaled 335 
in ArcGIS to real size and lichen thalli were outlined manually through visual inspection of the 336 
photos and measured automatically with high accuracy according to the diameter of the smallest 337 
circle which can circumscribe the lichen outline (Suppl. Fig. SF1). We preferred the simple 338 
geometrical shape of the circle and its diameter to identify the largest axis to circumvent the 339 
problem of complex-shaped lichens. Similar procedures for lichen thalli measurement from 340 
photographs are outlined in Hooker and Brown (1977). 341 

Then, we initially applied a 0.44 mm yr−1 constant growth rate and a 10-year colonization lag 342 
(Kugelmann, 1991) to the measurement of the largest Rhizocarpon geographicum lichen (longest 343 
axis). This growth rate is derived from the lichen growth curve with the highest number of control 344 
points so far in Tröllaskagi (Kugelmann, 1991), and it is very similar to that reported from the 345 
near Hörgárdalur valley (Häberle, 1991). However, the authors are aware that using a constant 346 
growth rate implies not taking account the growth rate decline with increasing age. Other longer 347 
colonization lags of 15, 20, 25 and 30 years were added to the age estimate from the growth rate 348 
in order to test the colonization lag originally assumed by Kugelmann (1991), on the suspicion of 349 
longer colonization lags reported elsewhere (Caseldine, 1983; Evans et al., 1999, Table 1). The 350 
resulting ages were compared to the dates of historical aerial photographs where the glacier snout 351 
positions constrain the maximum and minimum ages of the lichen stations. 352 

In the case of Porpidia cf. soredizodes, since no growth rate value has been described so far, a 353 
value will be tentatively estimated in this study. For this reason, we took measurements of the two 354 
species in the same sampling locations wherever possible. It should be noted that visual distinction 355 
between Porpidia cf. soredizodes and Porpidia tuberculosa is not always conclusive based on 356 



morphological characteristics, but we feel confident using the measurements of the Porpidia cf. 357 
soredizodes we identified in the field. 358 

3.4. 36Cl Cosmic-Ray Exposure (CRE) dating 359 

Where the thalli either coalesced and prevented identification of the largest thallus or dating 360 
results indicated that a moraine was too old to be dated by this method, rock samples were 361 
collected for CRE dating. The criteria for boulder and surface selection where the same as for 362 
lichenometric purposes: stable boulders with no signs of being affected by slope processes 363 
(landslides, debris-flows) or postglacial overturning, well embedded in the moraine, and with no 364 
sign of surface weathering or previous boulder break. The cosmogenic nuclide 36Cl was chosen 365 
because of the basalt lithology ubiquitous in Iceland, which lacks quartz, which is needed for 366 
standard 10Be CRE methods. Using a hammer and chisel, samples were collected from flat-topped 367 
surfaces of moraine boulders. In order to obtain a maximum time constraint for the deglaciation 368 
of both surveyed valleys, two samples were taken from Elliði (Fig. 1), a glacially polished ridge 369 
downstream from the Tungnahryggsjökull forelands separating Kolbeinsdalur to the north and 370 
Víðinesdalur to the south. The laboratory procedures applied for 36Cl extraction from silicate 371 
whole rock samples were those described in Schimmelpfennig et al. (2011). Note that the samples 372 
had not enough minerals to perform the 36Cl extraction on mineral separates, which is generally 373 
the preferred approach to minimize the uncertainties in age exposure estimates, as in mineral 374 
separates 36Cl is often produced by less and better known production pathways than in whole rock 375 
samples (Schimmelpfennig et al., 2009). The samples were crushed and sieved to 0.25–1 mm in 376 
the Physical Geography Laboratory in the Complutense University, Madrid. Chemical processing 377 
leading to 36Cl extraction from the whole rock was carried out at the Laboratoire National des 378 
Nucléides Cosmogéniques (LN2C) at the Centre Europeén de Recherche et d'Enseignement des 379 
Géosciences de l'Environnement (CEREGE), Aix-en-Provence (France). Initial weights of about 380 
120 g per sample were used. A chemically untreated split of each sample was set aside for analyses 381 
of the chemical composition of the bulk rocks at CRPG-SARM. First, the samples were rinsed to 382 
remove dust and fines. Then, 25–30% of the mass was dissolved to remove atmospheric 36Cl and 383 
potentially Cl-rich groundmass by leaching with a mixture of ultra-pure dilute nitric (10% HNO3) 384 
and concentrated hydrofluoric (HF) acids. In the next step, 2 g aliquots were taken to determine 385 
the major element concentrations; these were analysed by ICP-OES at CRPG-SARM. Then, 386 
before total dissolution, ~260 μl of a 35Cl carrier solution (spike) manufactured in-house 387 
(concentration: 6.92 mg g−1; 35Cl/37Cl ratio: 917) were added to the sample for isotope dilution 388 
(Ivy-Ochs et al., 2004). Total dissolution was achieved with access quantities of the above 389 
mentioned acid mixture. Following total dissolution, the samples were centrifuged to remove the 390 
undissolved residues and gel (fluoride complexes such as CaF2). Next, chlorine was precipitated 391 
to silver chloride (AgCl) by adding 2 ml of silver nitrate (AgNO3) solution at 10%. After storing 392 
the samples for two days in a dark place to allow the AgCl to settle on the bottom, the supernatant 393 
acid solution was extracted by a peristaltic pump. To reduce the isobaric interferences of 36S 394 
during the 36Cl AMS measurements, the first precipitate was re-dissolved in 2 ml of ammonia 395 
(NH3 + H2O 1:1 vol → NH4OH), and 1 ml of a saturated solution of barium nitrate (Ba(NO3)2) 396 
was added to the samples to precipitate barium sulphate (BaSO4). It was removed by centrifuging 397 
and filtering the supernatant with a syringe through acrodisc filters. AgCl was precipitated again 398 
with 3–4 ml of diluted HNO3 (1:1 vol). The precipitate was collected by centrifuging, rinsed, and 399 
dried in an oven at 80 °C for 2 days. 400 

The final AgCl targets were analysed by accelerator mass spectrometry (AMS) to measure the 401 
35Cl/37Cl and 36Cl/35Cl ratios, from which the Cl and 36Cl concentration were inferred. The 402 



measurements were carried out at the Accélérateur pour les Sciences de la Terre, Environnement 403 
et Risques (ASTER) at CEREGE in March 2017 using inhouse standard SM-CL-12 with an 404 
assigned value of 1.428 (±0.021) × 10−12 for the 36Cl/35Cl ratio (Merchel et al., 2011) and assuming 405 
a natural ratio of 3.127 for the stable ratio 35Cl/37Cl. 406 

When calculating exposure ages, the Excel™ spreadsheet for in situ 36Cl exposure age 407 
calculations proposed by Schimmelpfennig et al. (2009) was preferred to other online calculators 408 
(e.g. CRONUS Earth; Marrero et al., 2016) as it allows input of different 36Cl production rates 409 
from spallation, referenced to sea level and high latitude (SLHL). Thus, three SLHL 36Cl 410 
production rates from Ca spallation, i.e. the most dominant 36Cl production reaction in our 411 
samples, were applied to allow comparisons both with other Icelandic areas (57.3 ± 5.2 atoms 36Cl 412 
(g Ca)−1 yr−1; Licciardi et al., 2008) and other areas of the world (48.8 ± 3.4 atoms 36Cl (g Ca)−1 413 
yr−1, Stone et al., 1996; 42.2 ± 4.8 atoms 36Cl (g Ca)−1 yr−1, Schimmelpfennig et al., 2011). For 414 
36Cl production reactions other than Ca spallation, the following SLHL 36Cl production 415 
parameters were applied: 148.1 ± 7.8 atoms 36Cl (g K)−1 yr−1 for K spallation (Schimmelpfennig 416 
et al., 2014b), 13 ± 3 atoms 36Cl (g Ti)−1 yr−1 for Ti spallation (Fink et al., 2000), 1.9 ± 0.2 atoms 417 
36Cl (g Fe)−1 yr−1 for Fe spallation (Stone et al., 2005) and 696 ± 185 neutrons (g air)−1 yr−1 for the 418 
production rate of epithermal neutrons for fast neutrons in the atmosphere at the land/atmosphere 419 
interface (Marrero et al., 2016). Elevation-latitude scaling factors were based on the time invariant 420 
“St” scheme (Stone, 2000). The high-energy neutron attenuation length value applied was 421 
160 g cm−2. 422 

All production rates from spallation of Ca mentioned above are based on calibration samples with 423 
a predominant Ca composition. Iceland is permanently affected by a low-pressure cell, the 424 
Icelandic Low (Einarsson, 1984). As the atmospheric pressure modifies the cosmic-ray particle 425 
flux, and thus has an impact on the local cosmogenic nuclide production rate, the atmospheric 426 
pressure anomaly has to be taken into account when scaling the SLHL production rates to the 427 
study site. Only Licciardi et al.'s (2008) production rate already accounts for this anomaly, as the 428 
calibration sites of study are located in south western Iceland (see also Licciardi et al., 2006). On 429 
the other hand, Stone et al.'s (1996) and Schimmelpfennig et al.'s (2011) production rates were 430 
calibrated in Tabernacle Hill (Utah, western U.S.A.) and Etna volcano (Italy), respectively, hence 431 
they need to be corrected for the atmospheric pressure anomaly when applied in Iceland. Dunai 432 
(2010) advises including any long-term atmospheric pressure anomaly at least for Holocene 433 
exposure periods. Therefore, the local atmospheric pressure at the sample locations was applied 434 
in the scaling factor calculations when using the Stone et al. (1996) and Schimmelpfennig et al. 435 
(2011) production rates. Instead of the standard value of 1013.25 hPa at sea level, a sea level value 436 
of 1006.9 hPa (Akureyri meteorological station; Icelandic Meteorological Office, 2018) was used. 437 
The atmospheric pressure correction assumed a linear variation of temperature with altitude. The 438 
results presented and discussed below are based on the 36Cl production rate for Ca spallation of 439 
Licciardi et al. (2008) as it is calibrated for Iceland and considers the atmospheric pressure 440 
anomaly. The exposure ages presented throughout the text and figures include analytical and 441 
production rate errors unless stated otherwise. Our <2000 yr CRE ages have also been rounded to 442 
the nearest decade, and then converted to CE dates through their subtraction from the year 2015 443 
(i.e. fieldwork and sampling campaign). In order to achieve robust comparisons of our results 444 
with those obtained by radiocarbon or tephrochronology, we have calibrated the 14C ages 445 
previously published in the literature through the OxCal 4.3 online calculator 446 
(https://c14.arch.ox.ac.uk/oxcal/OxCal.html) implementing the IntCal13 calibration curve 447 
(Reimer et al., 2013). 448 



4. Results 449 

4.1. Geomorphological mapping, aerial photos and identified glacial stages 450 

Based on photo-interpretation of aerial photographs and fieldwork, two geomorphological maps 451 
were generated at ~1:7000 scale, in which well-preserved moraine segments, current glacier 452 
margins and stream network were also mapped (Fig. 2, Fig. 3). In Vesturdalur, over 1000 moraine 453 
ridge fragments (including terminal and lateral moraine segments) were identified and mapped, 454 
with increased presence from 1.7 km upwards to the current glacier terminus. In the analysis, we 455 
retained only ridge fragments if they are at least 50 m long, protruding 2 m above the valley 456 
bottom and the alignment of glacial boulders embedded in the moraine crest is preserved, as 457 
indicators of major glacial culminations and well preservation state. Otherwise we considered 458 
either they represented insignificant glacial stages or were most probably affected by post-glacial 459 
slope reworking. 460 

Following these criteria, we retained 12 glacial stages based on the geomorphological mapping 461 
of the Vesturdalur foreland (Fig. 2). We were able to verify in the field that the selected sections 462 
had not been affected by postglacial slope processes as none of these moraines was cut by debris-463 
flows or deformated/covered by landslides. The palaeo-position of the glacier terminus was 464 
clearly defined by pairs of latero-frontal moraines in stages 1, 4 and 6. In stages 2, 3, 5, 9, 10, 11 465 
and 12, it was poorly defined by short latero-frontal moraines on one side of the valley, very close 466 
to the river, with their prolongation and intersection with the river assumed to be the former apex, 467 
and so a tentative terminus geometry was drawn. The lateral geometry of the tongue was 468 
accurately reconstructed in stages 6, 8 and 9 based on long and aligned, ridge fragments, and in 469 
the other stages an approximate geometry was drawn from the terminus to the headwall. The 470 
greatest retreat between consecutive stages (1 km) occurred in the transition from the stage 1 to 471 
the stage 2. No intermediate frontal moraines were observed in the transect between the stages 1 472 
and 2. 473 

In Austurdalur, over 1600 moraine ridge fragments were identified and mapped. Following the 474 
same criteria as in Vesturdalur, 13 stages were identified based on the geomorphological evidence 475 
(Fig. 3). The furthest moraines marking the maximum extent of the eastern Tungnahryggsjökull 476 
appear at 1.3 km from the current terminus. In contrast to the other valley, moraines populate the 477 
glacier foreland more densely and regularly on both sides of the valley. Most of them are <50 m 478 
long with a few exceeding 100–200 m. The frontal moraine ridge fragments in Austurdalur clearly 479 
represent the geometry of the terminus in all the stages, as they are well-preserved and are only 480 
bisected by the glacier meltwater, with the counterparts easily identifiable. The most prominent 481 
(over 2 m protruding over the bottom of the valley) and well preserved moraines are those 482 
marking the terminus position at stages 1 to 8, with lengths ranging from 170 to 380 m. 483 

Both glaciers were also outlined on aerial photos of 1946, 1985, 1994 (only for western 484 
Tungnahryggsjökull), and 2000, and on a SPOT satellite image (2005), and hence new recent 485 
stages from the past century were studied for the Tungnahryggsjökull, five for the western glacier 486 
(stages 13, 14, 15, 16 and 17) and four for the eastern glacier (stages 11, 15, 16 and 17) (Suppl. 487 
Fig. SF2; Suppl. Table ST1). The stages identified on the basis of the geomorphological mapping 488 
and those obtained from glacier outlining over historical aerial photos sum up to a total of 17 489 
stages for each glacier. 490 

4.2. Glacier length, extent and volume 491 



The reconstructed glacier surfaces corresponding to the different glacial stages are shown in 492 
Suppl. Fig. SF3. The length of the glaciers during their reconstructed maximum ice extent was 493 
unequal, with the western Tungnahryggsjökull being 6.5 km long, and the eastern glacier being 494 
3.8 km long (Suppl. Table ST2). The same occurred with the area, 9.4 and 5.3 km2, respectively 495 
(Suppl. Table ST3). Over the different stages, the western glacier lost 31% of its area and retreated 496 
51% of its total length (ST2, ST3) while there was less shrinkage of the eastern glacier both in 497 
area loss (26%) and length (34%). Figs. 2B and 3B and Suppl. Table ST2 show only one reversal 498 
during the general retreating trend, in the stage 15 of the western and the stage 16 of the eastern 499 
Tungnahryggsjökull. The greatest area losses occurred in the transition to the stages 2, 4, 7, 9 and 500 
14 (>3%) on the western glacier while losses between stages in the eastern glacier were lower 501 
except for the transition between the stages 14 and 17, where a noticeable reduction of the 502 
accumulation area was observed in the aerial photos of 1946 and 1985, and the satellite image of 503 
2005 (Fig. 3B). The volumes calculated from the reconstructed glacier DEM and the corrected 504 
bed DEM show that the glaciers reached ~1.10 km3 (western) and ~0.47 km3 (eastern) at their 505 
recorded maximum extent corresponding to their outmost moraines. From the oldest to the most 506 
recent stage they lost 30% and 23% of their ice volume, respectively. The losses between 507 
consecutive stages of the glaciers were in general lower than 3% with the exception of the stages 508 
2, 4, 7, 9 (western) and stage 14 (eastern), where the values ranged from 3% to 10% (Suppl. Table 509 
ST3). Only one slight inversion of the volume trend is seen in the stages 15 of the western glacier 510 
and in 16 of the eastern glacier. 511 

4.3. Equilibrium-Line Altitudes (ELAs) 512 

The application of the AAR (0.67) method showed ELAs ranging from 1021 to 1099 m a.s.l. 513 
(western glacier) and from 1032 to 1065 m a.s.l. (eastern glacier) for the different reconstructed 514 
stages. This means an overall rise of 78 and 33 m respectively (Table 1) from the maximum to the 515 
minimum extent recorded. The results from applying the AABR (1.5) show the same trend and 516 
similar ELAs, with differences of up to ±15 m compared with those obtained from the AAR. The 517 
AABR-ELAs tend to be higher than the AAR-ELAs, especially in the second half of the 518 
reconstructed stages with the most remarkable differences occurring in the eastern 519 
Tungnahryggsjökull (Table 1). However, the error derived from the uncertainties associated to the 520 
applied balance ratio (BR, ±0.4) tends to decrease as the glaciers get smaller. The greatest change 521 
in the ELA between consecutive stages is found between the stages 1 and 2 in the western glacier 522 
(+26 m), fully coincident with the largest retreat measured, about 1 km. An interesting result is 523 
that the ELA rise is attenuated in both glaciers from stage 10 onwards, with stage-to-stage 524 
variations close to zero predominating, and with only one inversion (−3 m) occurring in the 525 
western glacier between the stages 14 and 15 (Table 1). 526 

4.4. Lichenometric dating 527 

Altogether 17 lichenometry stations (8 in Vesturdalur and 9 in Austurdalur) were set up during 528 
fieldwork in the two glacier forelands. Their spatial distribution and correspondence with the 529 
different glacial stages are given in Fig. 2, Fig. 3 and Suppl. Table ST4. The table also shows the 530 
measurements of the largest Rhizocarpon geographicum thalli found during fieldwork in the 531 
moraines (stages) to which the stations correspond. TUE-0 was the only station where no lichen 532 
of either species was found in 2015. Unlike Porpidia cf. soredizodes thalli, which are present in 533 
all the remaining lichenometry stations, Rhizocarpon geographicum thalli suitable for measuring 534 
(ellipsoidal, not coalescent) were only found in stations TUW-2 to TUW-7 in Vesturdalur, and in 535 
TUE-2 to TUE-7 in Austurdalur. 536 



The measurements considered the diameter of the smallest circle bounding the thallus outline as 537 
representative of the longest axis (Suppl. Fig. SF1). The obtained values ranged from 19.3 to 538 
71.5 mm in Vesturdalur, and from 19.0 to 47.7 mm in Austurdalur (Suppl. Table ST4). Only one 539 
size inversion (decreasing size with increasing distance to the current terminus) was detected in 540 
Austurdalur in the station TUE-6 (Fig. 3). Rhizocarpon geographicum thalli were absent in the 541 
nearest stations (TUW-1 and TUE-1) and were coalescent in the most distant stations (TUW-8 542 
and TUE-8) in both glacier forelands. On the other hand, Porpidia cf. soredizodes thalli 543 
measurements ranged from 18.3 to 148.4 mm in Vesturdalur, and from 16.8 to 141.8 mm in 544 
Austurdalur (Suppl. Table ST4). A size inversion of Porpidia cf. soredizodes thalli is also 545 
observed in TUE-6 and in TUE-7. 546 

When the ages of Rhizocarpon geographicum lichens are calculated applying a 0.44 mm yr−1 547 
growth rate and a 10-year colonization lag following Kugelmann (1991), they range from 54 to 548 
173 years in Vesturdalur, and from 53 to 119 years in Austurdalur. If longer colonization lags of 549 
15, 20, 25 and 30 years (see Methods section) are applied tentatively, the oldest ages obtained are 550 
~170–190 years, and the youngest ~40–70 years (Table 2). In general, the further away the 551 
lichenometry stations are, the older are the ages (Fig. 2, Fig. 3), with the inversions mentioned 552 
above. Apparently the lichenometry-dated moraines are younger than the CRE-dated distal 553 
moraines. 554 

The absence of Rhizocarpon geographicum lichens in 2015 at lichenometry station TUW-1 555 
(uncovered by the glacier at some time between 1994 and 2000; Suppl. Fig. SF2) suggests a 556 
colonization lag of at least 15–21 years. In addition, it is only when the colonization lag is assumed 557 
to be longer than 10 years and shorter than 30 years that the ages obtained at stations TUW-3 and 558 
TUE-2 are in good agreement with the ages deduced from the aerial photos (Table 2). 559 

During the 2015 field campaign, the species Porpidia cf. soredizodes was absent in station TUE-560 
0, located on a glacially polished threshold uncovered by the glacier after 2005 (post-stage 17). 561 
However, it was found in stations TUW-1 and TUE-1, dated to 1994–2000 and 1946–1985 562 
respectively, based on the position of the snouts in the aerial photos (Suppl. Fig. SF2). These 563 
observations suggest a colonization lag from 10 to 15–21 years, thus shorter than for Rhizocarpon 564 
geographicum. 565 

4.5. 36Cl CRE dating 566 

The detailed geomorphological analysis carried out on the numerous moraine ridges of both 567 
valleys has greatly limited the number of boulders reliable for successfully applying 36Cl CRE 568 
and lichenometry dating methods. Due to the great intensity of the slope processes (especially 569 
snow avalanches and debris-flows), only a few boulders are well-preserved, sometimes only one 570 
in each moraine ridge that retains its original glacier location. Thus, it should be highlighted that 571 
this issue prevented performing a statistically valid sampling for both methods (see e.g. Schaefer 572 
et al., 2009; Heyman et al., 2011). 573 

During the fieldwork campaigns in the forelands of the western and eastern Tungnahryggsjökull 574 
glaciers, 12 samples from stable and very protrusive moraine boulders (Suppl. Fig. SF4) were 575 
collected in areas where lichenometric dating was not suitable because of lichen thalli 576 
coalescence, and also 2 samples from a polished ridge downwards the confluence of Vesturdalur 577 
and Austurdalur (Suppl. Fig. SF5). The input data for exposure age calculations, namely sample 578 
thickness, topographic shielding factor, major element concentrations of bulk/target fractions, are 579 
summarized in the ST5, ST6, ST7, ST8 includes the 36Cl CRE ages calculated according to 580 



different Ca spallation production rates and the distance to the most recent glacier terminus 581 
position mapped. Table 3 includes the 36Cl CRE ages converted to CE dates format presented 582 
throughout the text. The dates presented below are based on the Licciardi et al. (2008) Ca 583 
spallation production rate. 584 

Aiming to obtain a maximum (oldest) age for the onset of the deglaciation in the valleys studied, 585 
two samples (ELLID-1 and ELLID-2) were extracted from the western sector of Elliði, a 660-m-586 
high glacially polished crest separating Viðinesdalur and Kolbeinsdalur valleys (Suppl. Fig. SF5), 587 
and located at 11 km downstream from the confluence of Vesturdalur and Austurdalur valleys. 588 
Both samples yielded dates of 14,300 ± 1700 BCE and 14,200 ± 1700 BCE. 589 

In Vesturdalur, 8 samples were collected from 5 moraines corresponding to 5 of the stages 590 
identified on the geomorphological map (Fig. 2). Samples TUW-9 and TUW-10 were taken from 591 
two boulders in the moraine that correspond to the left latero-frontal edge of the glacier during 592 
the stage 7, ~1 km from the 2005 CE (stage 17) glacier terminus (measured along the flowline 593 
from the reconstructed terminus apex); they yielded consistent dates of 1590 ± 100 CE and 594 
1640 ± 90 CE. Samples TUW-11 and TUW-12 were extracted from the moraines corresponding 595 
to the glacial stages 6 and 5, at 185 m and 465 m downstream respectively (~1.2 km and ~1.5 km 596 
respectively from the 2005 CE snout), and gave consistent dates 1480 ± 120 CE and 597 
1450 ± 100 CE. The next 36Cl samples (TUW-13 and TUW-14) were taken from two adjacent 598 
moraine boulders on the left latero-frontal moraine ridge corresponding to the stage 3, at 264 m 599 
downstream (1.8 km from the 2005 CE snout); they yielded dates of 1470 ± 130 CE and 600 
670 ± 210 CE, which are significantly different from each other. The most distant samples (TUW-601 
15 and TUW-16) were extracted from the stage 2 moraine, ~400 m downstream from the moraine 602 
ridge corresponding to stage 3 (2.2 km from the 2005 CE terminus). They yielded dates of 603 
1460 ± 110 CE and 1220 ± 190 CE, respectively, that are consistent which each other and in 604 
chronological order with sample TUW-13 from the stage 3. However, these ages are not in 605 
agreement with the oldest date of TUW-14 (670 ± 210 CE) from the stage 3. 606 

Four 36Cl samples were taken from two prominent moraines in the Austurdalur valley (Fig. 3). 607 
Samples TUE-9 and TUE-10 were collected on the frontal moraine corresponding to stage 4 (1 km 608 
from the 2005 CE glacier terminus) and yield significantly different dates of 740 ± 170 CE and 609 
1460 ± 100 CE. Samples TUE-11 and TUE-12 were taken from two moraine boulders located on 610 
the ridge of the left lateral moraine that records the maximum ice extent (Fig. 3). Their calculated 611 
36Cl dates, 380 ± 200 CE and 400 ± 200 CE, are consistent with each other and in stratigraphic 612 
order with the ages from stage 4. 613 

The dates derived from Stone et al. (1996) Ca spallation production rate are similar to those 614 
presented above, only 4% older. Those derived from the Schimmelpfennig et al. (2011) production 615 
rate yielded dates older by 15% on average (Suppl. Table ST8). These small differences do not 616 
represent statistical difference given the calculated age uncertainties. 617 

5. Discussion 618 

5.1. 36Cl CRE dating 619 

The dates obtained in both valleys range from 380 ± 200 CE to 1640 ± 90 CE and are significantly 620 
younger than the ages from Elliði polished ridge (Suppl. Table ST8). TUW-14 (670 ± 210 CE) 621 
could be the only outlier as it is significantly older than the other sample obtained from the same 622 
moraine of the stage 3 (TUW-13; 1470 ± 130 CE) and also older than the samples TUW-15 623 



(1460 ± 110 CE) and TUW-16 (1220 ± 190 CE), from the moraine of the previous stage 2 (Fig. 624 
2). This would imply assuming nuclide inheritance for sample TUW-14 due to either 625 
remobilization of an earlier exposed moraine boulder (see Matthews et al., 2017) or to previous 626 
exposure periods, but the high geomorphological dynamism of the slopes limits this possibility 627 
(Andrés et al., 2019). Another possible interpretation is that the samples TUW-15 and TUW-16 628 
(stage 2) would be the outliers since they may have experienced incomplete exposure due to post-629 
depositional shielding (Heyman et al., 2011). The possibility of both samples being affected by 630 
proglacial processes can be ruled out given the distance to the meltwaters channel or the absence 631 
of glacial burst features (see Caseldine, 1985a) in the surroundings of the sampled boulders. Thus, 632 
the ages of samples TUW-14 and TUW-13 would indicate that the moraine of stage 3 was built 633 
during two overlapped glacial advances at 670 ± 210 CE and 1470 ± 130 CE. In Austurdalur, the 634 
interpretation of the ages is also complex. Samples TUE-9 (740 ± 170 CE) and TUE-10 635 
(1460 ± 100 CE) from the same moraine are significantly different, but it is difficult to decide 636 
whether or not one of these two samples is an outlier. It is also possible that the younger age 637 
indicates the timing of a further readvance with the snout reaching the same moraine. It would be 638 
necessary to take more samples, but it will be difficult to find other sectors of these moraines that 639 
are not affected by slope processes, especially debris-flows. The hypothesis of supraglacial debris 640 
dumping can be rejected due to the lack of supraglacial debris or other features indicative of a 641 
palaeo debris-covered glacier. Other possibility to be considered is that older boulders may be 642 
incorporated in the formation of push moraines as has been shown in maritime Scandinavia, and 643 
identified to give overestimated ages for LIA-moraines (Matthews et al., 2017). The 644 
asynchronicity of glacial advances and retreat in both valleys should not be surprising as glaciers 645 
can retreat or advance differently in adjacent valleys due to a number of factors such as 646 
hypsometry, aspect, gradient, etc. Caseldine (1985b) suggested a high climate sensitivity of the 647 
glacier in Vesturdalur due to its steeper wally floor and thinness, which could explain a different 648 
behaviour of the glacier compared to the eastern glacier. 649 

5.1.1. Pre-LIA glacial advances 650 

36Cl CRE dating results from glacially polished surfaces on the Elliði crest show an age of 651 
14,250 ± 1700 yr (mean) for the Kolbeinsdalur deglaciation (Suppl. Fig. SF5). However, this 652 
should be considered as a minimum age since this probably indicates when the retreating and 653 
thinning glacier uncovered the ridge, and thus the start of the final deglaciation of the main valley. 654 
From this point to the outmost frontal moraines surveyed in this paper, no glacially polished 655 
outcrops or erratic boulders suitable for 36Cl sampling were found, impeding us to provide further 656 
chronological constraints for the deglaciation pattern of these valleys. 657 

Our 36Cl CRE dating results suggest late Holocene glacial advances prior to the LIA, at around 658 
~400 and ~700 CE, in Vesturdalur and Austurdalur, respectively, coinciding with the Dark Ages 659 
Cold Period (DACP) (between 400 and 765 CE) in central Europe, according to Helama et al. 660 
(2017). In fact, recent synthesis about Icelandic lake records indicates a strong decline in 661 
temperature at 500 CE (Geirsdóttir et al., 2018). 662 

The presence of Late Holocene moraines outside the outermost LIA moraines in other valleys of 663 
Tröllaskagi has been suggested through radiocarbon dating and tephrochronology in the 664 
Vatnsdalur (first, between 4880 ± 325 cal. BCE (tephra Hekla 5) and 3430 ± 510 cal. BCE, and 665 
another after 1850 ± 425 cal. BCE; Stötter, 1991), Lambárdalur (before 3915 ± 135 cal. BCE; 666 
Wastl and Stötter, 2005), Þverárdalur (before 3500 ± 130 cal. BCE; Wastl and Stötter, 2005), 667 
Kóngsstaðadalur (after 1945 ± 170 cal. BCE; Wastl and Stötter, 2005), Barkárdalur (between 668 



375 ± 375 cal. BCE and 190 ± 340 cal. CE and before 460 ± 200 cal. CE; Häberle, 1991) and 669 
Bægisárdalur (2280 ± 205 cal. BCE and 1050 ± 160 cal. CE; Häberle, 1991) valleys. Specifically, 670 
our oldest 36Cl dates (~400 and ~700 CE) coincide with glacial advances that are radiocarbon 671 
dated from intra-morainic peat bogs in other valleys in Tröllaskagi and southern Iceland, e.g. the 672 
Barkárdalur II stage (ca. 460 CE) in Subatlantic times (Häberle, 1991, Häberle, 1994), and given 673 
the uncertainties of our dates (200 yr), also with Drangajökull (NW Iceland) advancing at the 674 
same time (~300 CE; Harning et al., 2018). On the other hand, Meyer and Venzke (1985) had 675 
already suggested the presence of pre-LIA moraines at Klængshóll (eastern cirque, tributary of 676 
Skíðadalur). Caseldine, 1987, Caseldine, 1991 proposed a date of 5310 ± 345 cal. BCE as the 677 
youngest for a moraine in that cirque, based on tephrochronology (tephra layer Hekla 5 dated in 678 
a similar ground in Gljúfurárdalur) and rock weathering measurements using the Schmidt hammer 679 
technique. Based on the big size of some moraines in Skíðadalur and Holárdalur, Caseldine, 1987, 680 
Caseldine, 1991 also supports the hypothesis of pre-LIA moraines formed during several 681 
advances. All this information is not in agreement with the hypothesis that all the glaciers in 682 
Tröllaskagi reached their maximum ice extent since the Early Holocene during the LIA (Hjort et 683 
al., 1985; Caseldine, 1987, Caseldine, 1991; Caseldine and Hatton, 1994). 684 

In south and central and northwest Iceland, the ice caps reached their Late Holocene maximum 685 
advance during the LIA (Brynjólfsson et al., 2015a; Larsen et al., 2015; Harning et al., 2016a; 686 
Anderson et al., 2018; Geirsdóttir et al., 2018). But also pre-LIA advances have been identified 687 
and dated through radiocarbon and tephrochronology in a moraine sequence of Kvíárjökull 688 
(south-east Vatnajökull), the first of which occurred at Subatlantic times (before 110 ± 240 cal. 689 
BCE) and the second at 720 ± 395 CE (Black, 1990; cited in Guðmundsson, 1997). Considering 690 
the uncertainty of our ~400 and ~700 CE dates, these advances likely were coetaneous with other 691 
advances reported from southern Iceland in Kötlujökull (after 450 ± 100 cal. CE, radiocarbon 692 
dated and supported by tephochronology; Schomacker et al., 2003) and Sólheimajökull (southern 693 
Mýrkdalsjökull, “Ystagil stage”; Dugmore, 1989). In fact, the 669 ± 211 CE advance of western 694 
Tungnahryggsjökull (stage 3) also overlaps with the Drangajökull and Langjökull advances at 695 
~560 CE and 550 CE, respectively (Larsen et al., 2011; Harning et al., 2016a) during DACP, in 696 
response to the summer cooling between ~250 CE and ~750 CE (Harning et al., 2016a). These 697 
advances would correspond to the general atmospheric cooling in the North Atlantic reflected by 698 
widespread glacier advances (Solomina et al., 2016). 699 

The timing of the stage 2 in Vesturdalur is difficult to define because the sample TUW-16 700 
(1220 ± 190 CE) has almost 200-year uncertainty, and thus overlaps with both the LIA and the 701 
higher temperatures of the Medieval Warm Period (MWP, Lamb, 1965; 950–1250 CE, see 702 
Solomina et al., 2016) not conducive to glacier development as it has been observed in northwest 703 
and central Iceland (Larsen et al., 2011; Harning et al., 2016a). On the other hand, moraines in 704 
Greenland Arctic environments date within the MWP, based on cosmogenic nuclide dating 705 
(Young et al., 2015; Jomelli et al., 2016), suggesting a cooling in the western North Atlantic while 706 
the eastern sector remained warm. Iceland is located in the middle of this dipole “see-saw” pattern 707 
(Rogers and van Loon, 1979). Thus, the correlations with the glacier fluctuations Greenland or 708 
Norway, are complicated as the climate anomalies in both regions show opposing signs in the 709 
different see-saw modes. To increase the difficulty of interpretation, glaciers of Tröllaskagi are 710 
known to surge occasionally (Brynjólfsson et al., 2012; Ingólfsson et al., 2016), which could 711 
explain additional complexity in the glacial advances pattern, as it would not be driven by climatic 712 
variability. Brynjólfsson et al. (2012) pointed out that only four surges from three glaciers 713 



(Teigarjökull, Búrfellsjökull, Bægisárjökull) have been reported in the Tröllaskagi peninsula, 714 
where over 160 cirque glaciers exist. 715 

5.1.2. LIA glacial advances (15th–17th centuries) 716 

The remaining 36Cl ages obtained from Vesturdalur and Austurdalur are within the 1450–1640 CE 717 
range (Suppl. Table ST8; Table 3) and so correspond to different glacial advances or standstills 718 
during the second third of the LIA during the 15th, 16th and 17th centuries. According to our results, 719 
one of the largest glacier extents of the LIA culminated in both valleys at the latest around the 720 
mid-15th century (samples: TUW-12: 1450 ± 100 CE/stage 6; TUE-10: 1460 ± 100 CE/stage 4). 721 
This is a very early date compared to the LIA advances previously dated in Tröllaskagi. Earlier 722 
lichenometric research carried out in nearby valleys report more recent dates for the LIA 723 
maximum: mid-18th century in Barkárdalur (Häberle, 1991); early 19th century in Bægisárjökull 724 
and Skríðudalur (Häberle, 1991), and in Búrfellsdalur and Vatnsdalur (Kugelmann, 1991, Fig. 7); 725 
mid-19th century (1845–1875 CE) in Myrkárjökull, Vindheimajökull (Häberle, 1991), 726 
Þverárdalur, Teigardalur, Grýtudalur, Vesturárdalur (Kugelmann, 1991, Fig. 7), 727 
Heiðinnamannadalur and Kvarnárdalur (Caseldine, 1991); and late 19th century (1880s–1890s 728 
CE) in Gljúfurárdalur (Caseldine, 1985b, Caseldine, 1991). Caseldine (1991) proposed 1810–729 
1820 CE as the date of the “outer” moraine of Vesturdalur based on a minimum lichenometric 730 
age. According to our 36Cl moraine ages, the LIA maximum of Tungnahryggsjökull glaciers 731 
occurred ~400 years earlier than these dates. This should not be surprising as Kirkbride and 732 
Dugmore (2001) reported lichenometric ages >100 years younger than those derived from 733 
tephrochronology in the same landforms. Our results also give the earliest LIA dates based on 734 
moraine dating obtained so far in Iceland, compared to the previously published dates all obtained 735 
through lichenometric dating: 1740–1760 CE in south-east Iceland (Chenet et al., 2010). In 736 
northwestern Iceland, glaciolacustrine sediments recorded a contemporary expansion of 737 
Drangajökull at ~1400 CE (Harning et al., 2016a). Likewise a first major advance of Langjökull 738 
(western central Iceland; Fig. 1A) between 1450 and 1550 CE has been reported (Larsen et al., 739 
2011), as well as advances in central Iceland between 1690 and 1740 CE (Kirkbride and Dugmore, 740 
2006), based on varve thickness variance together with annual layer counting and C:N mass ratio, 741 
and tephrochronology (geochemical analysis), respectively. 742 

The dates of samples TUW-13, TUW-12 and TUW-11 in Vesturdalur (average date 743 
1470 ± 120 CE; stages 3, 5 and 6) cannot be distinguished statistically if their internal uncertainty 744 
is considered (Fig. 4; Suppl. Table ST8). The spatial scatter of these sampling sites (Fig. 2) may 745 
indicate frequent and intense terminus fluctuations in a short time interval. Different ages obtained 746 
from the moraines of the stages 3 and 4 in Vesturdalur and Austurdalur, respectively, may indicate 747 
that in the 15th century the glacier termini reached the moraines deposited at ~700 CE (Fig. 2, Fig. 748 
3) and rebuilt them, which may explain the large size of these moraines. Caseldine (1987) argued 749 
that other moraines in Skíðadalur were formed during more than one advance. Based on the size 750 
of the largest moraines, he also pointed out that many glaciers must have advanced to positions 751 
similar to those of the late 19th century earlier in the Neoglacial. 752 

The glacier advances in the 15th century in Vesturdalur and Austurdalur may have been the result 753 
of different climate forcings during that century: negative radiative forcing and summer cooling 754 
were linked to intense volcanic activity (Miller et al., 2012), the low solar activity of the Spörer 755 
Minimum (1460–1550 CE; Eddy, 1976), the sea-ice/ocean feedback with increased sea ice (Miller 756 
et al., 2012) and the weakening of the Atlantic Meridional Overturning Circulation (AMOC) 757 
(Zhong et al., 2011). Solar activity may have played a major role in these glacial advances if we 758 



consider that the climate in the North Atlantic is highly influenced by solar activity variability 759 
(Jiang et al., 2015). On the contrary, historical records, although not determinant, suggest a mild 760 
climate for 1430–1550 CE (Ogilvie and Jónsdóttir, 2000), and especially for 1412–1470 CE 761 
(Ogilvie, 1984). This would also be compatible with increased precipitation and hence more 762 
winter snow in warm periods (Caseldine and Stotter, 1993; Stötter et al., 1999; Fernández-763 
Fernández et al., 2017), and also with potential surge activity with glacial advances not directly 764 
linked to specific climate periods (e.g. Brynjólfsson et al., 2012; Ingólfsson et al., 2016). 765 

In Vesturdalur, the dates of samples TUW-9 (1590 ± 100 CE) and TUW-10 (1640 ± 90 CE) are 766 
compatible with a cold period from the late 1500s to 1630 CE (Ogilvie and Jónsson, 2001). 767 
Nevertheless, considering the uncertainties of these samples, their dates also overlap with the 768 
Maunder Minimum (1645–1715 CE) (Eddy, 1976), when the LIA maximum glacial advance 769 
occurred in the Alps (Holzhauser et al., 2005). 770 

5.2. Lichenometric dating 771 

5.2.1. 36Cl CRE dating vs. lichenometric dating. Multiple lichen species dating approach 772 

The contrast between our CRE dating results and earlier lichenometry-based results published 773 
elsewhere evidences the clear underestimation of the latter. This could be explained either by 774 
lichen growth inhibition due to saturation of the rock surface and competition of other thalli (Wiles 775 
et al., 2010; Le Roy et al., 2017) or a colonization lag longer than assumed up to now, from 10 to 776 
15 years in Tröllaskagi (Kugelmann, 1991; Caseldine, 1985b, respectively). However, other 777 
factors affect the reliability of the lichen-derived ages, and may explain the differences with CRE 778 
dates. One of them is the reliability of the growth rates and lichenometric calibration curves that 779 
are commonly assumed to be linear (constant) growth in northern Iceland (e.g. Caseldine, 1983, 780 
Caseldine, 1985b; Häberle, 1991; Kugelmann, 1991), although it has been demonstrated that the 781 
lichen diameter growth declines with age (see e.g. Winkler, 2003, Fig. 4). This might lead to 782 
significant age underestimations for the oldest dates, putting potentially earlier dates in the 19th 783 
century. Most of the fixed control points from which lichen growth rates have been derived in 784 
northern Iceland comprise abandoned gravestones, memorial stones, old bridges and mostly 785 
abandoned farmsteads (Caseldine, 1983; Kugelmann, 1991). Time of death is commonly assumed 786 
for gravestones, and abandonment date for farmsteads. Although the latter is well known on the 787 
basis of historical documentation, the colonization lag sometimes relies on the “the likely 788 
duration” of the deterioration of the buildings after abandonment that depends on the quality and 789 
stability of the constructions (see Kugelmann, 1991), and hence affect the results. However, it 790 
should be highlighted that our approach circumvents this last issue by combining field 791 
observations with historical aerial photographs. In addition, given that lichen growth of 792 
Rhizocarpon subgenus depends mostly on available humidity, the location of the measured lichens 793 
has also a potential effect on the results, i.e. to micro-climatic changes (see Innes, 1985; Hamilton 794 
and Whalley, 1995) between the location of the fixed points and the location on the moraine (crest: 795 
dryer and more exposed; green zone on the proximal slope, etc.). In the literature cited above 796 
about lichenometry-dated LIA advances no detailed information is provided about the location of 797 
the lichen measured for dating purposes, so potential error derived from this issue cannot be 798 
assessed. Nevertheless, given that we measured lichens on horizontal flat surfaces where there is 799 
no restriction to the moisture receipt of the lichens, we can consider our results presented 800 
throughout the next sections valid. In any case, such a great difference of our CRE dates and the 801 
previous lichen-derived is more likely to be explained by the technical limitations of the technique 802 
rather than environmental factors. 803 



Our field observations and historical aerial photos of known age restricted the dates of the 804 
moraines colonized by Rhizocarpon geographicum lichens. This approach would support the 805 
dates obtained from the 0.44 mm yr−1 growth rate of Kugelmann (1991) and the 20-year 806 
colonization lag. This rate is slightly higher than those reported from the Antarctic Peninsula 807 
(0.31 mm yr−1; Sancho et al., 2017), but lower than in Tierra del Fuego (0.63 mm yr−1; Sancho et 808 
al., 2011). The correlation between the sizes of largest Rhizocarpon geographicum and Porpidia 809 
cf. soredizodes thalli at the same stations showed a proportionality between the largest thalli of 810 
both species: thalli of Porpidia cf. soredizodes species grow faster than those of Rhizocarpon 811 
geographicum. The sizes of the largest Rhizocarpon geographicum thalli were plotted against the 812 
sizes of the largest Porpidia cf. soredizodes lichens (Fig. 5). The slope of the best-fit linear curve 813 
obtained was around 1.675 (r2 = 0.82) that suggests that Porpidia cf. soredizodes lichens grow 814 
1.675 times faster than those of Rhizocarpon geographicum. So if we assume 0.44 mm yr−1 815 
growth rate for Rhizocarpon geographicum lichens, a tentative growth rate of Porpidia cf. 816 
soredizodes would be 0.737 mm yr−1. However, this approach should be taken with caution due 817 
to the limited number of lichens (n = 7) used of each species. Aiming to obtain a tentative date 818 
(Table 2), this growth rate was applied to the largest Porpidia cf. soredizodes lichens found in 819 
those stations where no Rhizocarpon geographicum lichens could be measured (TUW-1, TUW-820 
8, TUE-1 and TUE-8). 821 

Assuming colonization lags of 20 and 15 years for Rhizocarpon geographicum and Porpidia cf. 822 
soredizodes lichens, respectively, and the above-mentioned growth rates, several glacial advances 823 
or standstills are tentatively placed in the 19th and 20th centuries, in the context of a general retreat 824 
from the most advanced LIA positions, as discussed in the next section. 825 

5.2.2. LIA glacial advances/standstills: 19th century 826 

Despite the low number of lichenometry stations surveyed and the limitations of this dating 827 
approach discussed in the previous sections, our results are in accordance with the 828 
geomorphological logic and with the chronological frame obtained from historical aerial photos. 829 
The ages derived from Rhizocarpon geographicum thalli measurements at lichenometry stations 830 
pre-dating 1946 suggest that the western and eastern Tungnahryggsjökull culminated successive 831 
advances/standstills around the 1830s, 1840s, 1860s and 1890s CE (Fig. 2, Fig. 3, Table 2). Ages 832 
tentatively estimated from Porpidia cf. soredizodes thalli considering its apparently higher growth 833 
rate suggest glacial advances/standstills around the 1800s (Fig. 2, Fig. 3, Table 2). No 834 
inconsistence (age inversions) is detected in the lichenometry-dated moraine sequence, but we 835 
recognise that some age underestimation may occur since: (i) we assume linear growth when 836 
applying a growth rate; and (ii) our 19th dates were derived from >40–45 mm-diameter lichens 837 
(Table 2 and Suppl. Table ST4), which exceed the size of the biggest control point from which 838 
Kugelmann's growth rate was estimated (Kugelmann, 1991, Fig. 3). This problem reinforces the 839 
need of taking lichen-derived ages as relative. The interpretation of the chronology in Austurdalur 840 
is more complicated: the lichen sizes measured at the stations TUE-6 (1890s CE, stage 9) and 841 
TUE-7 (1880s CE; stage 7) yield more recent dates than at station TUE-5 (1860s CE; stage 10) 842 
despite being representative of earlier stages (Fig. 3). Given that this anomaly occurs in both 843 
lichen species, it is reasonable to conclude that an environmental factor could have affected the 844 
lichen populations at both stations. It could even indicate a date when the boulders may have been 845 
remobilized by postglacial processes. 846 

Our chronology of the 19th century glacial advances in both valleys (Table 2) does not coincide 847 
exactly with the phases identified by Kugelmann (1991) in Svarfaðardalur-Skíðadalur (1810, 848 



1850, 1870–1880, 1890–1900 CE). It also differs appreciably from the chronology proposed by 849 
Caseldine (1985b) in Vesturdalur (moraines of 1868, 1878, 1887, 1898 CE). Nevertheless, it 850 
should be borne in mind that we are considering only two glaciers with their own glaciological 851 
properties. Moreover, we have estimated the ages from the growth rate (instead of a specific 852 
calibration growth curve), longer colonization lag based on field observations, and applied to the 853 
longest axis measurements (Kugelmann, 1991). On the other hand, micro-climatic differences 854 
between the location of fixed points and our lichen measured may occur. However, we recognise 855 
that the comparisons with Caseldine's (1985b) results are quite difficult since: (i) the moraines 856 
dated by Caseldine are poorly located in his mapping (and hard to identify over aerial photos and 857 
our moraine mapping); (ii) we are not able to apply the same parameters to his measurements, as 858 
he used the mean longest axis of the five largest lichen is not provided (only the average value); 859 
and (iii) he recognised the lichen growth slowdown when dating the outermost moraine, which 860 
probably is one of those dated with 36Cl in this work. 861 

Our results, despite the limitations of the applied method, largely due to the low number of 862 
moraine boulders suitable for its application can be considered compatible and in good agreement 863 
with glacial advances and cold periods during the first third of the 19th century as we show below. 864 
Martin et al. (1991) pointed out the existence of a moraine dated to ca. 1810–1820 CE in “Western 865 
Tröllaskagi Tungnahryggsjökull” (ambiguously, without providing any detail of its location), as 866 
well as others of this period in Svarfaðardalur, Búrfellsdalur and Vatnsdalur. Other dates obtained 867 
from more sophisticated statistical techniques applied to lichenometry dating procedures also 868 
evidenced a glacial advance phase 1810–1820 CE (Chenet et al., 2010). These advance phases 869 
were coetaneous with significant sea ice persistence in the early 1800s (Ogilvie and Jónsson, 870 
2001), decreased solar activity, and strong volcanic eruptions (Dalton Minimum, 1790–1830 CE; 871 
Wagner and Zorita, 2005). The moraines dated in Vesturdalur to around 1830 CE (TUW-7: 1830s 872 
CE, stage 10) and the early 1890s CE (TUW-5: 1890s CE, stage 12) are in good agreement with 873 
the glacial advances in Svarfaðardalur-Skíðadalur during the 1830s and early 1890s CE 874 
(Kugelmann, 1991, Fig. 8), as well as with the low temperatures and presence of sea ice at these 875 
dates (Koch, 1945; Ogilvie, 1996; Ogilvie and Jónsdóttir, 2000; Ogilvie and Jónsson, 2001; 876 
Kirkbride, 2002). 877 

5.2.3. Post-LIA glacial advances/standstills 878 

Lichen-derived ages from Rhizocarpon geographicum thalli suggest the occurrence of glacial 879 
advances or standstills at the first half of the 20th century, culminating in 1910s, 1930s, 1940s and 880 
1950s CE, although the 1930s CE date (TUE-3; stage 12) disagrees with the date inferred from 881 
the aerial photos, at some point between 1946 and 1985 CE (Suppl. Fig. SF2; Table 2). However, 882 
the 1910s CE date in both valleys is in good agreement with the moraine abandonment during the 883 
first two decades of the 20th century; it would be the result of the accumulated effect of the 884 
temperature rise since the latter half of the 19th century (Caseldine, 1987; Wanner et al., 2008). 885 
Thus, the subsequent advances would be the response to specific relative thermal minima within 886 
a warmer climate (Stötter et al., 1999). The date of TUW-3 (stage 13) representing the moraine 887 
abandonment in ~1940 CE is in agreement with the overall context of general glacier retreat as a 888 
result of the warmest decades of the 1930s and 1940s CE, which triggered an intense retreat of 889 
the glaciers (Einarsson, 1991; Martin et al., 1991; Kirkbride, 2002). The advances/standstills 890 
dated to the early 1950s CE in both Tungnahryggsjökull glaciers are likely to be synchronous 891 
given the similarity of the dates obtained (Table 2). They could represent glacial advances in 892 
consonance with the late 1940s – early/mid-1950s CE cooling (Einarsson, 1991; Fernández-893 
Fernández et al., 2017), recorded both in Akureyri and many other weather stations throughout 894 



Iceland. Caseldine (1983) found a similar chronology in Gljúfurárdalur (Skíðadalur headwater, 895 
10 km to the east), with moraine deposition between mid-1910s and 1930, around mid-1930s and 896 
late 1940s-1950 CE. 897 

The subsequent trend of the Tungnahryggsjökull glaciers, inferred from aerial photographs, a 898 
satellite image, geomorphological mapping, and glacier reconstruction, was characterized by 899 
continuous retreat and volume loss, in line with the increasing temperature trend since the end of 900 
the LIA. This trend was only reversed between the mid-1960s and mid-1980s CE by a major 901 
cooling event (Einarsson, 1991; Sigurðsson, 2005; Fernández-Fernández et al., 2017). Two 902 
moraines have been dated after the 1950s CE. The date obtained in station TUE-1 (1970s CE; 903 
stage 14) agrees with the date deduced from the 1946 and 1985 aerial photographs (Suppl. Fig. 904 
SF2). However, the date estimated in station TUW-1 (1970s CE; between stages 15 and 16) is 905 
prior to that obtained from the 1994 and 2000 aerial photographs. The reason for this mismatch 906 
may be a non-linear growth phase of the Porpidia cf. soredizodes thalli measured (i.e. its real 907 
growth rate may have been lower than the estimated). Nevertheless, more research on this species 908 
is required to use it successfully in lichenometric dating. 909 

The aerial photographs from 1994 (stage 15 western Tungnahryggsjökull) and 2000 (stage 16) 910 
show a reversal in the trend of both Tungnahryggsjökull glaciers (Suppl. Fig. SF2), as the 911 
positions of their termini are more advanced than in 1985 (stages 14 and 15; see Fernández-912 
Fernández et al., 2017). These advances in Vesturdalur and Austurdalur, culminating after 1985 913 
in a period non-conducive to glacier expansion, may have been linked to above average 914 
precipitation for 1988–1995, which prevented a negative mass balance of the glaciers 915 
(Sigurðsson, 2005). The 2000 aerial photo (stage 16) and 2005 SPOT satellite image (stage 17) 916 
display the retreat of both glaciers. This trend was reinforced by the sudden warming initiated in 917 
1995, which triggered decreased snowfall, negative mass balances for 1996–2000, and retreat of 918 
the non-surging glaciers after 2000 (Sigurðsson, 2005). In spite of these glacial fluctuations in 919 
response to the intense climatic fluctuations of the last century, the ELAs estimated in this paper 920 
only show a general rise of 5–10 m (Table 1). This small ELA increase may be derived from some 921 
artifacts of the glacier reconstruction or the attenuating effect of the increase in winter 922 
precipitation suggested by glacier-climate models (Caseldine and Stotter, 1993; Fernández-923 
Fernández et al., 2017). 924 

5.3. Final remarks: can the occurrence of pre-LIA glacial advances be confirmed? Was the 925 
LIA in northern Iceland the Holocene glacial maximum? Was it a single maximum advance? 926 

Our results demonstrate that Tungnahryggsjökull advanced during Late Holocene glacial stages 927 
prior to the LIA and reached considerably more advanced positions, even though the Iceland large 928 
ice caps reached generally their Late Holocene maximum advance during the LIA (Larsen et al., 929 
2015; Harning et al., 2016a; Anderson et al., 2018; Geirsdóttir et al., 2018). The chronological 930 
data presented in the previous sections suggests that during the LIA the glaciers overlapped 931 
moraines deposited in pre-LIA glacial stages. Our results also show LIA advances since the 15th 932 
century and are thus contrary to the traditional proposal of a single maximum LIA advance 933 
occurring during either mid-18th or late-19th centuries in Tröllaskagi peninsula, followed by 934 
subsequent minor readvances in an overall retreating trend (see Caseldine, 1983, Caseldine, 935 
1985b, Caseldine, 1987; Kugelmann, 1991; Martin et al., 1991). The use of the 36Cl production 936 
rates from Ca spallation reported by Stone et al. (1996) does not lead to major changes in the 937 
interpretation and conclusions as the changes in the nominal dates (up to 90 yr earlier; Table 3) 938 
are within the external uncertainty and also overlap with the results presented above based on the 939 



Licciardi et al. (2008) production rate. Similarly, if we consider the results derived by the 940 
Schimmelpfennig et al. (2011) production rate, we obtain even earlier ages, which still overlap 941 
with the results derived by the other production rates due to a higher external uncertainty (see 942 
Table 3). Although the snow cover duration is known to be high in northern Iceland (see Dietz et 943 
al., 2012), quantifying the effect of snow cover on sub-surface 36Cl production is a complex issue: 944 
on the one side, snow lowers the isotopic production rates related to spallation reactions due to 945 
the shielding effect on high-energy neutrons (Benson et al., 2004; Schildgen et al., 2005), and on 946 
the other side it increases the 36Cl production rate from thermal neutrons below the rock surface, 947 
due to the enhancing effect of hydrogen on these low-energy neutrons (Dunai et al., 2014). Both 948 
effects might cancel out depending on the composition of the samples, and their quantification is 949 
still debated and affected by high uncertainties (Zweck et al., 2013; Dunai et al., 2014; Delunel et 950 
al., 2014). Given this complexity, snow shielding was not applied. In any case its effect is unlikely 951 
to be higher than the uncertainty derived from extracting 36Cl from whole rock instead of minerals. 952 

Our results are in concordance with several valleys in northen Iceland, but also with the results 953 
that are being obtained in other sectors of Arctic and North Atlantic region in the last years, such 954 
as: in Baffin Island (northern Canada), where glaciers reached Late Holocene maximum positions 955 
prior to the LIA (Young et al., 2015; 10Be dating), or West Greenland, with several advances or 956 
stabilizations at 1450 ± 90 CE and 1720 ± 60 CE (Jomelli et al., 2016; 36Cl dating). However, it is 957 
striking that the early LIA advances of northern Iceland reported in the present paper do not agree 958 
with maritime Norway with later maximum culminations (see Nesje et al., 2008) despite being 959 
the glaciated region most comparable with Iceland in terms of climate and glaciology. Our results 960 
also agree with farther and southern areas such as the Alps, with maximum advances at around 961 
1430 CE (Schimmelpfennig et al., 2012, Schimmelpfennig et al., 2014a); 10Be dates; and Sierra 962 
Nevada (Iberian Peninsula), where LIA advances have been reported between the 14th and 17th 963 
centuries (Palacios et al., 2019; 10Be dates). Our detailed moraine mapping, combined with CRE 964 
dating in the surveyed valleys, clearly shows a number of advances throughout the LIA, in 965 
response to the great climatic variability of the region with alternating cold and mild/warm periods 966 
(Ogilvie, 1984, Ogilvie, 1996; Ogilvie and Jónsdóttir, 2000; Ogilvie and Jónsson, 2001; 967 
Geirsdóttir et al., 2009) as occurred in the Alps (e.g. Schimmelpfennig et al., 2014a) or the Iberian 968 
mountains (e.g. Oliva et al., 2018). The ELA calculations show that the major long-lasting rise of 969 
the glacier ELA (24–50 m depending on the calculation method) took place prior to 1900s/1910s. 970 

According to the results, the evolution pattern described by Fernández-Fernández et al. (2017) for 971 
the Tungnahryggsjökull glaciers should be revised as the 36Cl CRE dates from the 15th and 17th 972 
centuries indicate that LIA maximum was reached earlier than previously thought. Thus, if we 973 
consider the ELA of the earliest LIA dates obtained with 36Cl CRE (i.e. stages 5 and 4 of western 974 
and eastern Tungnahryggsjökull), these imply ELA depressions (with reference to the 2005 date) 975 
of 24–50 m (depending on the ELA calculation method), occurring for at least 560 years (Table 976 
1) and not 150 years as had been previously assumed (Caseldine and Stotter, 1993; Fernández-977 
Fernández et al., 2017). 978 

6. Conclusions 979 

(i) This paper highlights the detailed geomorphological analysis of the glacial landforms as 980 
an essential pre-requisite prior to the application of dating methods on them. 981 
Nevertheless, this has greatly limited the number of valid moraine boulders to be dated, 982 
since the vast majority of those were affected by post-glacial processes. This issue which 983 
has prevented a statistically acceptable sampling and the validation or invalidation of the 984 



“inconsistent” ages yielded by several boulders, so a conclusive explanation cannot be 985 
given. 986 

(ii) Applying 36Cl CRE dating for the first time in the Tröllaskagi peninsula enabled us to 987 
identify pre-LIA glacial advances in Vesturdalur and Austurdalur. Thus, the western and 988 
eastern Tungnahryggsjökull glaciers did not reach their Late Holocene maximum extent 989 
during the LIA. The maximum extent for the eastern glacier was dated to ~400 CE. For 990 
the western glacier a latest date of ~700 CE and an earliest of 16,300 years ago (when the 991 
Elliði crest was deglaciated) have been obtained. 992 

(iii) The LIA maximum in Vesturdalur and Austurdalur was reached by the 15th century at the 993 
latest. A combination of detailed moraine mapping and 36Cl CRE dating confirm a number 994 
of glacial advances between the 15th and 17th centuries, the earliest LIA advances dated 995 
in Tröllaskagi at present. 996 

(iv) For the recent dates, the complementary use of aerial photographs, a satellite image and 997 
fieldwork has aided to obtain lichenometry-derived ages tentatively in good agreement 998 
with the morpho-stratigraphic order of the glacial landforms. It has also compensated 999 
some of the limitations and error sources of lichenometric dating. Thus, it has proved to 1000 
be a useful tool to assess the colonization lags and the validity of the lichenometry-derived 1001 
ages. 1002 

(v) Fieldwork on recently deglaciated surfaces and historical aerial photographs have shown 1003 
clearly that the colonization lag of Rhizocarpon geographicum lichen species is from 15–1004 
21 to 30 years, considerably longer than previously assumed in Tröllaskagi. Colonization 1005 
of the Porpidia cf. soredizodes species is shorter, between 10 and 21 years. 1006 

(vi) From the measurements carried out in different lichenometry stations, growth of Porpidia 1007 
cf. soredizodes lichen appeared to be higher than in the case of Rhizocarpon 1008 
geographicum, and also proportional to that. Its growth rate has been tentatively estimated 1009 
for the first time, at around 0.737 mm yr−1. However, further research on the growth rates 1010 
of this species is required for its potential use in lichenometric dating as a complementary 1011 
species, since it also shows a shorter colonization lag than Rhizocarpon geographicum. 1012 

Acknowledgements 1013 

This paper was supported by the project CGL2015-65813-R (Spanish Ministry of Economy and 1014 
Competitiveness) and Nils Mobility Program (EEA Grants), and with the help of the High 1015 
Mountain Physical Geography Research Group (Complutense University of Madrid). We thank 1016 
the Icelandic Association for Search and Rescue, the Icelandic Institute of Natural History, the 1017 
Hólar University College, and David Palacios Jr. and María Palacios for their support in the field. 1018 
José M. Fernández-Fernández received a PhD fellowship from the FPU programme (Spanish 1019 
Ministry of Education, Culture and Sport; reference FPU14/06150). The 36Cl measurements were 1020 
performed at the ASTER AMS national facility (CEREGE, Aix en Provence), which is supported 1021 
by the INSU/CNRS and the ANR through the “Projets thématiques d'excellence” program for the 1022 
“Equipements d'excellence” ASTER-CEREGE action and IRD. 1023 

  1024 



References 1025 

Anderson, L.S., Flowers, G.E., Jarosch, A.H., Aðalgeirsdóttir, G.T., Geirsdóttir, Á., Miller, G.H., 1026 
Harning, D.J., Thorsteinsson, T., Magnússon, E., Pálsson, F., 2018. Holocene glacier and climate 1027 
variations in Vestfirðir, Iceland, from the modeling of Drangajökull ice cap. Quat. Sci. Rev. 190, 1028 
39–56. https://doi.org/10.1016/j.quascirev.2018.04.024. 1029 

Andrés, N., Tanarro, L.M., Fernández, J.M., Palacios, D., 2016. The origin of glacial alpine 1030 
landscape in Tröllaskagi Peninsula (North Iceland). Cuad. Investig. Geogr. 42, 341–368. 1031 
https://doi.org/10.18172/cig.2935. 1032 

Andrés, N., Palacios, D., Sæmundsson, Þ., Brynjólfsson, S., Fernández-Fernández, J.M., 2019. 1033 
The rapid deglaciation of the Skagafjörður fjord, northern Iceland. Boreas 1034 
https://doi.org/10.1111/bor.12341. 1035 

Andrews, J.T., Giraudeau, J., 2003.Multi-proxy records showing significant Holocene 1036 
environmental variability: the inner N. Iceland shelf (Húnaflói). Quat. Sci. Rev. 22, 175–193. 1037 
https://doi.org/10.1016/S0277-3791(02)00035-5. 1038 

Arróniz-Crespo, M., Pérez-Ortega, S., De Los Ríos, A., Green, T.G.A., Ochoa-Hueso, R., 1039 
Casermeiro, M.Á., De La Cruz, M.T., Pintado, A., Palacios, D., Rozzi, R., Tysklind, N., Sancho, 1040 
L.G., 2014. Bryophyte-cyanobacteria associations during primary succession in recently 1041 
deglaciated areas of Tierra del Fuego (Chile). PLoS One 9, 15–17. 1042 
https://doi.org/10.1371/journal.pone.0096081. 1043 

Barker, S., Knorr, G., Vautravers, M.J., Diz, P., Skinner, L.C., 2010. Extreme deepening of the 1044 
Atlantic overturning circulation during deglaciation. Nat. Geosci. 3, 567–571. 1045 
https://doi.org/10.1038/ngeo921. 1046 

Benn, D.I., Hulton, N.R.J., 2010. An Excel™ spreadsheet program for reconstructing the surface 1047 
profile of former mountain glaciers and ice caps. Comput. Geosci. 36, 605–610. 1048 
https://doi.org/10.1016/j.cageo.2009.09.016. 1049 

Benson, L., Madole, R., Phillips,W., Landis, G., Thomas, T., Kubik, P., 2004. The probable 1050 
importance of snow and sediment shielding on cosmogenic ages of north-central Colorado 1051 
Pinedale and pre-Pinedale moraines. Quat. Sci. Rev. 23, 193–206. 1052 
https://doi.org/10.1016/j.quascirev.2003.07.002. 1053 

Bergþórsson, P., 1969. An estimate of drift ice and temperature in Iceland in 1000 years. Jökull 1054 
19, 94–101. 1055 

Bickerton, R.W., Matthews, J.A., 1992. On the accuracy of lichenometric dates: an assessment 1056 
based on the “Little Ice Age” moraine sequence of Nigardsbreen, southern Norway. The Holocene 1057 
2, 227–237. https://doi.org/10.1177/095968369200200304. 1058 

Björnsson, H., 1978. Surface area of glaciers in Iceland. Jökull 28, 31. 1059 

Black, T.A., 1990. The Holocene Fluctuation of the Kvíárjökull Glacier, Southeastern Iceland. 1060 
(Unpublished MA thesis). University of Colorado. 1061 

Bradwell, T., 2001. A new lichenometric dating curve for Southeast Iceland. Geogr. Ann. Ser. A 1062 
Phys. Geogr. 83, 91–101. https://doi.org/10.1111/j.0435-3676.2001.00146.x. 1063 

https://doi.org/10.1016/j.quascirev.2018.04.024
https://doi.org/10.18172/cig.2935
https://doi.org/10.1111/bor.12341
https://doi.org/10.1016/S0277-3791(02)00035-5
https://doi.org/10.1371/journal.pone.0096081
https://doi.org/10.1038/ngeo921
https://doi.org/10.1016/j.cageo.2009.09.016
https://doi.org/10.1016/j.quascirev.2003.07.002
https://doi.org/10.1177/095968369200200304
https://doi.org/10.1111/j.0435-3676.2001.00146.x


Bradwell, T., 2004a. Lichenometric dating in southeast Iceland: the size-frequency approach. 1064 
Geogr. Ann. Ser. A Phys. Geogr. 86, 31–41. https://doi.org/10.1111/j.0435-3676.2004.00211.x. 1065 

Bradwell, T., 2004b. Annual moraines and summer temperatures at Lambatungnajökull, Iceland. 1066 
Arct. Antarct. Alp. Res. 36, 502–508. https://doi.org/10.1657/1523-1067 
0430(2004)036[0502:AMASTA]2.0.CO;2. 1068 

Bradwell, T., Armstrong, R.A., 2006. Growth rates of Rhizocarpon geographicum lichens: a 1069 
review with new data from Iceland. J. Quat. Sci. 22, 311–320. https://doi.org/10.1002/jqs.1058. 1070 

Brugger, K.A., 2006. Late Pleistocene climate inferred from the reconstruction of the Taylor River 1071 
glacier complex, southern Sawatch Range, Colorado. Geomorphology 75, 318–329. 1072 
https://doi.org/10.1016/j.geomorph.2005.07.020. 1073 

Brynjólfsson, S., Ingólfsson, Ó., Schomacker, A., 2012. Surge fingerprintings of cirque glaciers 1074 
at Tröllaskagi peninsula, North Iceland. Jökull 62, 153–168. 1075 

Brynjólfsson, S., Schomacker, A., Guðmundsdóttir, E.R., Ingólfsson, Ó., 2015a. A 300-year surge 1076 
history of the Drangajökull ice cap, northwest Iceland, and itsmaximum during the Little Ice Age. 1077 
The Holocene 25 (7), 1076–1092. https://doi.org/10.1177/0959683615576232. 1078 

Brynjólfsson, S., Schomacker, A., Ingólfsson, Ó., Keiding, J.K., 2015b. Cosmogenic 36Cl 1079 
exposure ages reveal a 9.3 ka BP glacier advance and the Late Weichselian-Early Holocene glacial 1080 
history of the Drangajökull region, northwest Iceland. Quat. Sci. Rev. 126, 140–157. 1081 
https://doi.org/10.1016/j.quascirev.2015.09.001. 1082 

Caseldine, C., 1983. Resurvey of the margins of Gljúfurárjökull and the chronology of recent 1083 
deglaciation. Jökull 33, 111–118. 1084 

Caseldine, C.J., 1985a. Survey of Gljúfurárjökull and features associated with a glacier burst in 1085 
Gljúfurárdalur, Northern Iceland. Jökull 35, 61–68. 1086 

Caseldine, C.J., 1985b. The extent of some glaciers in Northern Iceland during the little ice age 1087 
and the nature of recent deglaciation. Geogr. J. 151, 215–227. https://doi.org/10.2307/633535. 1088 

Caseldine, C., 1987. Neoglacial glacier variations in northern Iceland: examples from the 1089 
Eyjafjörður area. Arct. Alp. Res. 19, 296–304. 1090 

Caseldine, C.J., 1990. A review of dating methods and their application in the development of a 1091 
Holocene glacial chronology for Northern Iceland. Gletscher-Und Landschaftsgeschichtliche 1092 
Untersuchungen in Nordisland, pp. 59–82. 1093 

Caseldine, C.J., 1991. Lichenometric dating, lichen population studies and Holocene glacial 1094 
history in Tröllaskagi, Northern Iceland. In: Maizels, J.K., Caseldine, C. (Eds.), Environmental 1095 
Change in Iceland: Past and Present. Glaciology and Quaternary Geology vol. 7. Springer, 1096 
Dordrecht, pp. 219–233. https://doi.org/10.1007/978-94-011-3150-6_15. 1097 

Caseldine, C.J., Cullingofrd, R.A., 1981. Recent Mapping of Gljúfurárjökull and Gljúfurárdalur. 1098 
Jökull 31, 11–22. 1099 

Caseldine, C., Hatton, J., 1994. Environmental change in Iceland.  Munch. Geogr. Abh. Reihe B 1100 
12, 41–62. 1101 

https://doi.org/10.1111/j.0435-3676.2004.00211.x
https://doi.org/10.1657/1523-0430(2004)036%5b0502:AMASTA%5d2.0.CO;2
https://doi.org/10.1657/1523-0430(2004)036%5b0502:AMASTA%5d2.0.CO;2
https://doi.org/10.1002/jqs.1058
https://doi.org/10.1016/j.geomorph.2005.07.020
https://doi.org/10.1177/0959683615576232
https://doi.org/10.1016/j.quascirev.2015.09.001
https://doi.org/10.2307/633535
https://doi.org/10.1007/978-94-011-3150-6_15


Caseldine, C., Stotter, J., 1993. “Little Ice Age” glaciation of Tröllaskagi peninsula, northern 1102 
Iceland: climatic implications for reconstructed equilibrium line altitudes (ELAs). The Holocene 1103 
3, 357–366. https://doi.org/10.1177/095968369300300408. 1104 

Chen, T., Robinson, L.F., Burke, A., Southon, J., Spooner, P., Morris, P.J., Ng, H.C., 2015. 1105 
Synchronous centennial abrupt events in the ocean and atmosphere during the last deglaciation. 1106 
Science 349, 1537–1541. https://doi.org/10.1126/science.aac6159. 1107 

Chenet, M., Roussel, E., Jomelli, V., Grancher, D., 2010. Asynchronous Little Ice Age glacial 1108 
maximum extent in southeast Iceland. Geomorphology 114, 253–260. 1109 
https://doi.org/10.1016/j.geomorph.2009.07.012. 1110 

Coquin, J., Mercier, D., Bourgeois, O., Cossart, E., Decaulne, A., 2015. Gravitational spreading 1111 
of mountain ridges coeval with Late Weichselian deglaciation: impact on glacial landscapes in 1112 
Tröllaskagi, northern Iceland. Quat. Sci. Rev. 107, 197–213. 1113 
https://doi.org/10.1016/j.quascirev.2014.10.023. 1114 

Cossart, E., Mercier, D., Decaulne, A., Feuillet, T., Jónsson, H.P., Sæmundsson, Þ., 2014. Impacts 1115 
of post-glacial rebound on landslide spatial distribution at a regional scale in northern Iceland 1116 
(Skagafjörður). Earth Surf. Process. Landf. 39, 336–350. https://doi.org/10.1002/esp.3450. 1117 

Crochet, P., Jóhannesson, T., Jónsson, T., Sigurðsson, O., Björnsson, H., Pálsson, F., Barstad, I., 1118 
2007. Estimating the spatial distribution of precipitation in Iceland using a linear model of 1119 
orographic precipitation. J. Hydrometeorol. 8, 1285–1306. 1120 
https://doi.org/10.1175/2007JHM795.1. 1121 

Dahl, S.O., Nesje, A., 1992. Paleoclimatic implications based on equilibrium-line altitude 1122 
depressions of reconstructed Younger Dryas and Holocene cirque glaciers in inner Nordfjord, 1123 
western Norway. Palaeogeogr. Palaeoclimatol. Palaeoecol. 94, 87–97. 1124 
https://doi.org/10.1016/0031-0182(92)90114-K. 1125 

Decaulne, A., 2016. Lichenometry in Iceland, results and application. Géomorphol. Relief 1126 
Process. Environ. 22, 77–91. https://doi.org/10.4000/geomorphologie.11291. 1127 

Decaulne, A., Saemundsson, T., 2006. Geomorphic evidence for present-day snowavalanche and 1128 
debris-flow impact in the Icelandic Westfjords. Geomorphology 80, 80–93. 1129 
https://doi.org/10.1016/J.GEOMORPH.2005.09.007. 1130 

Delunel, R., Bourlès, D.L., van der Beek, P.A., Schlunegger, F., Leya, I., Masarik, J., Paquet, E., 1131 

2014. Snow shielding factors for cosmogenic nuclide dating inferred from long-term neutron 1132 
detector monitoring. Quat. Geochronol. 24, 16–26. 1133 
https://doi.org/10.1016/J.QUAGEO.2014.07.003. 1134 

Dietz, A.J., Wohner, C., Kuenzer, C., 2012. European snow cover characteristics between 2000 1135 
and 2011 derived from improved MODIS daily snow cover products. Remote Sens. 4, 2432–1136 
2454. https://doi.org/10.3390/rs4082432. 1137 

Dong, G., Zhou, W., Yi, C., Zhang, L., Li, M., Fu, Y., Zhang, Q., 2017. Cosmogenic 10Be surface 1138 
exposure dating of ‘Little Ice Age’ glacial events in the Mount Jaggang area, central Tibet. The 1139 
Holocene 27 (10), 1516–1525. https://doi.org/10.1177/0959683617693895. 1140 

https://doi.org/10.1126/science.aac6159
https://doi.org/10.1016/j.geomorph.2009.07.012
https://doi.org/10.1016/j.quascirev.2014.10.023
https://doi.org/10.1002/esp.3450
https://doi.org/10.1175/2007JHM795.1
https://doi.org/10.1016/0031-0182(92)90114-K
https://doi.org/10.4000/geomorphologie.11291
https://doi.org/10.1016/J.GEOMORPH.2005.09.007
https://doi.org/10.1016/J.QUAGEO.2014.07.003
https://doi.org/10.3390/rs4082432
https://doi.org/10.1177/0959683617693895


Dugmore, A.J., 1989. Tephrochronological studies of Holocene glacial fluctuations in South 1141 
Iceland. Glacier Fluctuations and Climatic Change, pp. 37–55 https://doi.org/10.1007/978-94-1142 
015-7823-3_3. 1143 

Dunai, T.J., 2010. Cosmogenic Nuclides. Cambridge University Press, Cambridge 1144 
https://doi.org/10.1017/CBO9780511804519. 1145 

Dunai, T.J., Binnie, S.A., Hein, A.S., Paling, S.M., 2014. The effects of a hydrogen-rich ground 1146 
cover on cosmogenic thermal neutrons: implications for exposure dating. Quat. Geochronol. 22, 1147 
183–191. https://doi.org/10.1016/J.QUAGEO.2013.01.001. 1148 

Eddy, J.A., 1976. The Maunder minimum. Science 192 (4245), 1189–1202. 1149 
https://doi.org/10.1126/science.192.4245.1189. 1150 

Einarsson, M.Á., 1984. Climate of Iceland. In: van Loon, H. (Ed.), World Survey of Climatology: 1151 
15: Climates of the Oceans. Elsevier, Amsterdam, pp. 673–697. 1152 

Einarsson, M.A., 1991. Temperature conditions in Iceland 1901–1990. Jökull 41, 1–20. 1153 

Etzelmüller, B., Farbrot, H., Guðmundsson, Á., Humlum, O., Tveito, O.E., Björnsson, H., 2007. 1154 
The regional distribution of mountain permafrost in Iceland. Permafr. Periglac. Process. 18, 185–1155 
199. https://doi.org/10.1002/ppp.583. 1156 

Evans, D.J.A., Archer, S., Wilson, D.J.H., 1999. A comparison of the lichenometric and Schmidt 1157 
hammer dating techniques based on data from the proglacial areas of some Icelandic glaciers. 1158 
Quat. Sci. Rev. 18, 13–41. https://doi.org/10.1016/S0277-3791(98)00098-5. 1159 

Fernández-Fernández, J.M., Andrés, N., 2018. Methodological proposal for the analysis of the 1160 
evolution of glaciers since the Little Ice Age and its application in the Tröllaskagi Peninsula 1161 
(Northern Iceland). Cuad. Investig. Geogr. Geogr. Res. Lett. 44, 69–97. 1162 
https://doi.org/10.18172/cig.3392. 1163 

Fernández-Fernández, J.M., Andrés, N., Sæmundsson, Þ., Brynjólfsson, S., Palacios, D., 2017. 1164 
High sensitivity of North Iceland (Tröllaskagi) debris-free glaciers to climatic change from the 1165 
‘Little Ice Age’ to the present. The Holocene 27, 1187–1200. 1166 
https://doi.org/10.1177/0959683616683262. 1167 

Feuillet, T., Coquin, J., Mercier, D., Cossart, E., Decaulne, A., Jónsson, H.P., Sæmundsson, Þ., 1168 
2014. Focusing on the spatial non-stationarity of landslide predisposing factors in northern 1169 
Iceland. Prog. Phys. Geogr. 38, 354–377. https://doi.org/10.1177/0309133314528944. 1170 

Fink, D., Vogt, S., Hotchkis, M., 2000. Cross-sections for 36Cl from Ti at Ep=35–150 MeV: 1171 
applications to in-situ exposure dating. Nucl. Instruments Methods Phys. Res. Sect. B Beam 1172 
Interact. with Mater. Atoms 172, 861–866. https://doi.org/10.1016/S0168-583X(00)00200-7. 1173 

Flowers, G.E., Björnsson, H., Geirsdóttir, Á., Miller, G.H., Black, J.L., Clarke, G.K.C., 2008. 1174 
Holocene climate conditions and glacier variation in central Iceland from physical modelling and 1175 
empirical evidence. Quat. Sci. Rev. 27, 797–813. 1176 
https://doi.org/10.1016/J.QUASCIREV.2007.12.004. 1177 

Gardner, A.S., Moholdt, G., Cogley, J.G., Wouters, B., Arendt, A.A., Wahr, J., Berthier, E., Hock, 1178 
R., Pfeffer, W.T., Kaser, G., Ligtenberg, S.R.M., Bolch, T., Sharp, M.J., Hagen, J.O., Van Den 1179 

https://doi.org/10.1007/978-94-015-7823-3_3
https://doi.org/10.1007/978-94-015-7823-3_3
https://doi.org/10.1017/CBO9780511804519
https://doi.org/10.1016/J.QUAGEO.2013.01.001
https://doi.org/10.1126/science.192.4245.1189
https://doi.org/10.1002/ppp.583
https://doi.org/10.1016/S0277-3791(98)00098-5
https://doi.org/10.18172/cig.3392
https://doi.org/10.1177/0959683616683262
https://doi.org/10.1177/0309133314528944
https://doi.org/10.1016/S0168-583X(00)00200-7
https://doi.org/10.1016/J.QUASCIREV.2007.12.004


Broeke, M.R., Paul, F., 2013. A reconciled estimate of glacier contributions to sea level rise: 2003 1180 
to 2009. Science 340, 852–857. https://doi.org/10.1126/science.1234532. 1181 

Geirsdóttir, Á., Miller, G.H., Axford, Y., Ólafsdóttir, Sædís, 2009. Holocene and latest Pleistocene 1182 
climate and glacier fluctuations in Iceland. Quat. Sci. Rev. 28, 2107–2118. 1183 
https://doi.org/10.1016/j.quascirev.2009.03.013. 1184 

Geirsdóttir, Á., Miller, G.H., Andrews, J.T., Harning, D.J., Anderson, L.S., Thordarson, T., 2018. 1185 
The onset of Neoglaciation in Iceland and the 4.2 ka event. Clim. Past Discuss. 1186 
https://doi.org/10.5194/cp-2018-130 (Manuscript under review). 1187 

Gordon, J.E., Sharp, M., 1983. Lichenometry in dating recent glacial landforms and deposits, 1188 
southeast Iceland. Boreas 12, 191–200. https://doi.org/10.1111/j.1502-3885.1983.tb00312.x. 1189 

Grove, J.M., 1988. The Little Ice Age. Routledge, London. 1190 

Guðmundsson, H.J., 1997. A review of the Holocene environmental history of Iceland. Quat. Sci. 1191 
Rev. 16, 81–92. https://doi.org/10.1016/S0277-3791(96)00043-1. 1192 

Häberle, T., 1991. Holocene glacial history of the Hörgárdalur area, Tröllaskagi, Northern Iceland. 1193 
In: Maizels, J.K., Caseldine, C. (Eds.), Environmental Change in Iceland: Past and Present. 1194 
Glaciology and Quaternary Geology vol. 7. Springer, Dordrecht, pp. 193–202. 1195 
https://doi.org/10.1007/978-94-011-3150-6_13. 1196 

Häberle, T., 1994. Glacial, Late Glacial and Holocene history of the Hörgárdalur area, 1197 

Tröllaskagi, Northern Iceland. In: Stöttter, J., Wilhelm, F. (Eds.), Environmental Change in 1198 
Iceland. Münchener Geographische Abhandlungen, Reihe B, pp. 133–145. 1199 

Hamilton, S.J.,Whalley,W.B., 1995. Rock glacier nomenclature: a re-assessment. 1200 
Geomorphology 14, 73–80. https://doi.org/10.1016/0169-555X(95)00036-5. 1201 

Harning, D.J., Geirsdóttir, Á., Miller, G.H., Anderson, L., 2016a. Episodic expansion of 1202 
Drangajökull, Vestfirðir, Iceland, over the last 3 ka culminating in its maximum dimension during 1203 
the Little Ice Age. Quat. Sci. Rev. 152, 118–131. https://doi.org/10.1016/j.quascirev.2016.10.001. 1204 

Harning, D.J., Geirsdóttir, Á., Miller, G.H., Zalzal, K., 2016b. Early Holocene deglaciation of 1205 
Drangajökull, Vestfirðir, Iceland. Quat. Sci. Rev. 153, 192–198. 1206 
https://doi.org/10.1016/j.quascirev.2016.09.030. 1207 

Harning, D.J., Geirsdóttir, Á., Miller, G.H., 2018. Punctuated Holocene climate of Vestfirðir, 1208 
Iceland, linked to internal/external variables and oceanographic conditions. Quat. Sci. Rev. 189, 1209 
31–42. https://doi.org/10.1016/j.quascirev.2018.04.009. 1210 

Harris, T., Tweed, F.S., Knudsen, Ó., 2004. A polygenetic landform at Stígá, Öræfajökull, 1211 
southern Iceland. Geogr. Ann. Ser. A Phys. Geogr. 86, 143–154. https://doi.org/10.1111/j.0435-1212 
3676.2004.00220.x. 1213 

Helama, S., Jones, P.D., Briffa, K.R., 2017. Dark Ages Cold Period: a literature review and 1214 
directions for future research. The Holocene 27, 1600–1606. 1215 
https://doi.org/10.1177/0959683617693898. 1216 

https://doi.org/10.1126/science.1234532
https://doi.org/10.1016/j.quascirev.2009.03.013
https://doi.org/10.5194/cp-2018-130
https://doi.org/10.1111/j.1502-3885.1983.tb00312.x
https://doi.org/10.1016/S0277-3791(96)00043-1
https://doi.org/10.1007/978-94-011-3150-6_13
https://doi.org/10.1016/0169-555X(95)00036-5
https://doi.org/10.1016/j.quascirev.2016.10.001
https://doi.org/10.1016/j.quascirev.2016.09.030
https://doi.org/10.1016/j.quascirev.2018.04.009
https://doi.org/10.1111/j.0435-3676.2004.00220.x
https://doi.org/10.1111/j.0435-3676.2004.00220.x
https://doi.org/10.1177/0959683617693898


Heyman, J., Stroeven, A.P., Harbor, J.M., Caffee, M.W., 2011. Too young or too old: evaluating 1217 
cosmogenic exposure dating based on an analysis of compiled boulder exposure ages. Earth 1218 
Planet. Sci. Lett. 302, 71–80. https://doi.org/10.1016/j.epsl.2010.11.040. 1219 

Hjort, C., Ingólfsson, Ó., Norðdahl, H., 1985. Late Quaternary geology and glacial history of 1220 
Hornstrandir, Northwest Iceland: a reconnaissance study. Jökull 35, 9–29. 1221 

Holzhauser, H., Magny, M., Zumbuühl, H.J., 2005. Glacier and lake-level variations in west 1222 
central Europe over the last 3500 years. The Holocene 15, 789–801. 1223 
https://doi.org/10.1191/0959683605hl853ra. 1224 

Hooker, T.N., Brown, D.H., 1977. A photographic method for accurately measuring the growth 1225 
of crustose and foliose saxicolous lichens. Lichenologist 9, 65–75. 1226 
https://doi.org/10.1017/S0024282977000073. 1227 

Hughes, P.D.,Woodward, J.C., van Calsteren, P.C., Thomas, L.E., Adamson, K.R., 2010. 1228 
Pleistocene ice caps on the coastal mountains of the Adriatic Sea. Quat. Sci. Rev. 29, 3690–3708. 1229 
https://doi.org/10.1016/j.quascirev.2010.06.032. 1230 

Icelandic Meteorological Office, 2018. Climatological data. Available. 1231 
http://en.vedur.is/climatology/data/, Accessed date: 13 April 2018. 1232 

Ingólfsson, Ó., Benediktsson, Í.Ö., Schomacker, A., Kjær, K., Brynjólfsson, S., Jónsson, S.A., 1233 
Korsgaard, N.J., Johnson, M., 2016. Glacial geological studies of surge type glaciers in Iceland – 1234 
research status and future challenges. Earth-Sci. Rev. 152, 37–69. 1235 

Innes, J.L., 1985. Lichenometry. Prog. Phys. Geogr. 1236 
https://doi.org/10.1177/030913338500900202. 1237 

Ipsen, H.A., Principato, S.M., Grube, R.E., Lee, J.F., 2018. Spatial analysis of cirques from three 1238 
regions of Iceland: implications for cirque formation and palaeoclimate. Boreas 47, 565–576. 1239 
https://doi.org/10.1111/bor.12295. 1240 

Ivy-Ochs, S., Synal, H.-A., Roth, C., Schaller, M., 2004. Initial results from isotope dilution for 1241 
Cl and 36Cl measurements at the PSI/ETH Zurich AMS facility. Nucl. Instrum. Methods Phys. 1242 
Res., Sect. B 223–224, 623–627. https://doi.org/10.1016/j.nimb.2004.04.115. 1243 

Jacob, T.,Wahr, J., Pfeffer,W.T., Swenson, S., 2012. Recent contributions of glaciers and ice caps 1244 
to sea level rise. Nature 482, 514–518. https://doi.org/10.1038/nature10847. 1245 

Jaksch, K., 1970. Beobachtungen in den Gletschervorfeldern des Sólheima und Siðujökull im 1246 
Sommer 1970. Jökull 20. 1247 

Jaksch, K., 1975. Das Gletschervorfeld des Sólheimajökull. Jökull 25, 34–38. 1248 

Jaksch, K., 1984. Das Gletcshervorfeld des Vatnajökull am Oberflauf des Djúpá, Südisland. Jökull 1249 
34, 97–103. 1250 

James, W.H.M., Carrivick, J.L., 2016. Automated modelling of spatially-distributed glacier ice 1251 
thickness and volume. Comput. Geosci. 92, 90–103. https://doi.org/10.1016/j.cageo.2016.04.007. 1252 

Janke, J.R., Bellisario, A.C., Ferrando, F.A., 2015. Classification of debris-covered glaciers and 1253 
rock glaciers in the Andes of central Chile. Geomorphology 241, 98–121. 1254 
https://doi.org/10.1016/j.geomorph.2015.03.034. 1255 

https://doi.org/10.1016/j.epsl.2010.11.040
https://doi.org/10.1191/0959683605hl853ra
https://doi.org/10.1017/S0024282977000073
https://doi.org/10.1016/j.quascirev.2010.06.032
http://en.vedur.is/climatology/data/
https://doi.org/10.1177/030913338500900202
https://doi.org/10.1111/bor.12295
https://doi.org/10.1016/j.nimb.2004.04.115
https://doi.org/10.1038/nature10847
https://doi.org/10.1016/j.cageo.2016.04.007
https://doi.org/10.1016/j.geomorph.2015.03.034


Jiang, H., Muscheler, R., Björck, S., Seidenkrantz,M.S., Olsen, J., Sha, L., Sjolte, J., Eiríksson, 1256 
J., Ran, L., Knudsen, K.L., Knudsen, M.F., 2015. Solar forcing of Holocene summer seasurface 1257 
temperatures in the northern North Atlantic. Geology 43, 203–206. 1258 
https://doi.org/10.1130/G36377.1. 1259 

Jóhannesson, H., Sæmundsson, K., 1989. Geological map of Iceland. 1:500.000. Bedrock. 1260 
Icelandic Institute of Natural History, Reykjavik. 1261 

Jóhannesson, T., Sigurðsson, O., 1998. Interpretation of glacier variations in Iceland 1930–1995. 1262 
Jökull 45, 27–33. 1263 

Jomelli, V., Lane, T., Favier, V., Masson-Delmotte, V., Swingedouw, D., Rinterknecht, V., 1264 
Schimmelpfennig, I., Brunstein, D., Verfaillie, D., Adamson, K., Leanni, L., Mokadem, F., 1265 
Aumaître, G., Bourlès, D.L., Keddadouche, K., 2016. Paradoxical cold conditions during the 1266 
medieval climate anomaly in the Western Arctic. Sci. Rep. 6, 32984. 1267 
https://doi.org/10.1038/srep32984. 1268 

Jónsson, O., 1976. Berghlaup. Ræktunarfélag Norðurlands, Akureyri. 1269 

Kirkbride, M.P., 2002. Icelandic climate and glacier fluctuations through the termination of the 1270 
“Little Ice Age”. Polar Geogr. 26, 116–133. https://doi.org/10.1080/789610134. 1271 

Kirkbride, M.P., 2011. Debris-covered glaciers. In: Singh, V.P., Singh, P., Haritashya, U.K. (Eds.), 1272 
Encyclopedia of Snow, Ice and Glaciers. Encyclopedia of Earth Series. Springer, Netherlands, pp. 1273 
180–182 https://doi.org/10.1007/978-90-481-2642-2_622. 1274 

Kirkbride,M.P., Dugmore, A.J., 2001. Can lichenometry be used to date the “Little Ice Age” 1275 
glacial maximum in Iceland? Clim. Chang. 48, 151–167. 1276 
https://doi.org/10.1023/A:1005654503481. 1277 

Kirkbride, M.P., Dugmore, A.J., 2006. Responses of mountain ice caps in central Iceland to 1278 
Holocene climate change. Quat. Sci. Rev. 25, 1692–1707. 1279 
https://doi.org/10.1016/j.quascirev.2005.12.004. 1280 

Knight, J., Harrison, S., Jones, D.B., 2018. Rock glaciers and the geomorphological evolution of 1281 
deglacierizing mountains. Geomorphology 311, 127–142. 1282 
https://doi.org/10.1016/j.geomorph.2018.09.020. 1283 

Koblet, T., Gärtner-Roer, I., Zemp,M., Jansson, P., Thee, P., Haeberli, W., Holmlund, P., 2010. 1284 
Reanalysis of multi-temporal aerial images of Storglaciären, Sweden (1959–99); part 1: 1285 
determination of length, area, and volume changes. Cryosphere 4, 333–343. 1286 
https://doi.org/10.5194/tc-4-333-2010. 1287 

Koch, L., 1945. The East Greenland ice. Medd. Grønland 130, 1–374. Kugelmann, O., 1991. 1288 
Dating recent glacier advances in the Svarfaðardalur-Skíðadalur area of Northern Iceland by 1289 
means of a new lichen curve. In: Maizels, J.K., Caseldine, C. (Eds.), Environmental Change in 1290 
Iceland: Past and Present. Glaciology and Quaternary Geology vol. 7. Springer, Dordrecht, pp. 1291 
203–217. https://doi.org/10.1007/978-94-011-3150-6_14. 1292 

Lamb, H.H., 1965. The early medieval warm epoch and its sequel. Palaeogeogr. Palaeoclimatol. 1293 
Palaeoecol. 1, 13–37. https://doi.org/10.1016/0031-0182(65)90004-0. 1294 

https://doi.org/10.1038/srep32984
https://doi.org/10.1080/789610134
https://doi.org/10.1007/978-90-481-2642-2_622
https://doi.org/10.1023/A:1005654503481
https://doi.org/10.1016/j.quascirev.2005.12.004
https://doi.org/10.1016/j.geomorph.2018.09.020
https://doi.org/10.5194/tc-4-333-2010
https://doi.org/10.1007/978-94-011-3150-6_14
https://doi.org/10.1016/0031-0182(65)90004-0


Larsen, D.J., Miller, G.H., Geirsdóttir, Á., Thordarson, T., 2011. A 3000-year varved record of 1295 
glacier activity and climate change from the proglacial lake Hvítárvatn, Iceland. Quat. Sci. Rev. 1296 
30, 2715–2731. https://doi.org/10.1016/j.quascirev.2011.05.026. 1297 

Larsen, D.J., Geirsdóttir, Á., Miller, G.H., 2015. Precise chronology of Little Ice Age expansion 1298 
and repetitive surges of Langjökull, central Iceland. Geology 43, 167–170. 1299 
https://doi.org/10.1130/G36185.1. 1300 

Le Roy, M., Deline, P., Carcaillet, J., Schimmelpfennig, I., Ermini, M., 2017. 10Be exposure 1301 
dating of the timing of Neoglacial glacier advances in the Ecrins-Pelvoux massif, southern French 1302 
Alps. Quat. Sci. Rev. 178, 118–138. https://doi.org/10.1016/j.quascirev.2017.10.010. 1303 

Li, Y., Li, Y., Harbor, J., Liu, G., Yi, C., Caffee, M.W., 2016. Cosmogenic10Be constraints on 1304 
Little Ice Age glacial advances in the eastern Tian Shan, China. Quat. Sci. Rev. 138, 105–118. 1305 
https://doi.org/10.1016/j.quascirev.2016.02.023. 1306 

Licciardi, J.M., Kurz,M.D., Curtice, J.M., 2006. Cosmogenic3He production rates from Holocene 1307 
lava flows in Iceland. Earth Planet. Sci. Lett. 246, 251–264. 1308 
https://doi.org/10.1016/j.epsl.2006.03.016. 1309 

Licciardi, J.M., Kurz, M.D., Curtice, J.M., 2007. Glacial and volcanic history of Icelandic table 1310 
mountains from cosmogenic 3He exposure ages. Quat. Sci. Rev. 26, 1529–1546. 1311 
https://doi.org/10.1016/j.quascirev.2007.02.016. 1312 

Licciardi, J.M., Denoncourt, C.L., Finkel, R.C., 2008. Cosmogenic36Cl production rates from Ca 1313 
spallation in Iceland. Earth Planet. Sci. Lett. 267, 365–377. 1314 
https://doi.org/10.1016/j.epsl.2007.11.036. 1315 

Maizels, J.K., Dugmore, A.J., 1985. Lichenometric dating and tephrochronology of sandur 1316 
deposits, Sólheimajökull area, southern Iceland. Jökull 35, 69–77. 1317 

Marrero, S.M., Phillips, F.M., Caffee, M.W., Gosse, J.C., 2016. CRONUS-Earth cosmogenic 36Cl 1318 
calibration. Quat. Geochronol. 31, 199–219. https://doi.org/10.1016/j.quageo.2015.10.002. 1319 

Martin, H.E., Whalley, W.B., Caseldine, C., 1991. Glacier fluctuations and rock glaciers in 1320 
Tröllaskagi, Northern Iceland, with special reference to 1946–1986. In: Maizels, J.K., Caseldine, 1321 
C. (Eds.), Environmental Change in Iceland: Past and Present. Springer Netherlands, Dordrecht, 1322 
pp. 255–265 https://doi.org/10.1007/978-94-011-3150-6_17. 1323 

Marzeion, B., Cogley, J.G., Richter, K., Parkes, D., 2014. Attribution of global glacier mass loss 1324 
to anthropogenic and natural causes. Science 345, 919–921. 1325 
https://doi.org/10.1126/science.1254702. 1326 

Matthews, J.A., Shakesby, R.A., Fabel, D., 2017. Very low inheritance in cosmogenic surface 1327 
exposure ages of glacial deposits: a field experiment from two Norwegian glacier forelands. The 1328 
Holocene 27, 1406–1414. https://doi.org/10.1177/0959683616687387. 1329 

Merchel, S., Bremser,W., Alfimov, V., Arnold, M., Aumaître, G., Benedetti, L., Bourlès, D.L., 1330 
Caffee, M., Fifield, L.K., Finkel, R.C., Freeman, S.P.H.T., Martschini, M., Matsushi, Y., Rood, 1331 
D.H., Sasa, K., Steier, P., Takahashi, T., Tamari, M., Tims, S.G., Tosaki, Y., Wilcken, K.M., Xu, 1332 
S., 2011. Ultra-trace analysis of 36Cl by acceleratormass spectrometry: an interlaboratory study. 1333 
Anal. Bioanal. Chem. https://doi.org/10.1007/s00216-011-4979-2. 1334 

https://doi.org/10.1016/j.quascirev.2011.05.026
https://doi.org/10.1130/G36185.1
https://doi.org/10.1016/j.quascirev.2017.10.010
https://doi.org/10.1016/j.quascirev.2016.02.023
https://doi.org/10.1016/j.epsl.2006.03.016
https://doi.org/10.1016/j.quascirev.2007.02.016
https://doi.org/10.1016/j.epsl.2007.11.036
https://doi.org/10.1016/j.quageo.2015.10.002
https://doi.org/10.1007/978-94-011-3150-6_17
https://doi.org/10.1126/science.1254702
https://doi.org/10.1177/0959683616687387
https://doi.org/10.1007/s00216-011-4979-2


Mercier, D., Cossart, E., Decaulne, A., Feuillet, T., Jónsson, H.P., Sæmundsson, Þ., 2013. The 1335 
Höfðahólar rock avalanche (sturzström): chronological constraint of paraglacial landsliding on an 1336 
Icelandic hillslope. The Holocene 23, 432–446. https://doi.org/10.1177/0959683612463104. 1337 

Meyer, H.H., Venzke, J.F., 1985. Der Klængshóll-Kargletscher in Nordisland. Nat. Mus. 115, 29–1338 
46. 1339 

Miller, G.H., Geirsdóttir, Á., Zhong, Y., Larsen, D.J., Otto-Bliesner, B.L., Holland,M.M., Bailey, 1340 
D.A., Refsnider, K.A., Lehman, S.J., Southon, J.R., Anderson, C., Björnsson, H., Thordarson, T., 1341 
2012. Abrupt onset of the Little Ice Age triggered by volcanism and sustained by sea-ice/ocean 1342 
feedbacks. Geophys. Res. Lett. 39, 1–5. https://doi.org/10.1029/2011GL050168. 1343 

National Land Survey of Iceland, 2018. Aerial photo collection. Available. 1344 
https://www.lmi.is/landupplysingar/loftmyndasafn-2-2/, Accessed date: 5 July 2018. 1345 

Nesje, A., Bakke, J., Dahl, S.O., Lie, Ø., Matthews, J.A., 2008. Norwegian mountain glaciers in 1346 
the past, present and future. Glob. Planet. Chang. 60, 10–27. 1347 
https://doi.org/10.1016/j.gloplacha.2006.08.004. 1348 

Ogilvie, A.E.J., 1984. The past climate and sea-ice record from Iceland, part 1: data to A.D. 1780. 1349 
Clim. Chang. 6, 131–152. https://doi.org/10.1007/BF00144609. 1350 

Ogilvie, A., 1996. Sea-ice Conditions off the Coasts of Iceland A. D. 1601–1850 With Special 1351 
Reference to Part of the Maunder Minimum Period (1675–1715). 25. AmS-Varia, pp. 9–12. 1352 

Ogilvie, A.E.J., Jónsdóttir, I., 2000. Sea ice, climate, and Icelandic fisheries in the eighteenth and 1353 
nineteenth centuries. Arctic 53, 383–394. https://doi.org/10.14430/arctic869. 1354 

Ogilvie, A.E.J., Jónsson, T., 2001. “Little Ice Age” research: a perspective from Iceland. Clim. 1355 
Chang. 48, 9–52. 1356 

Oliva, M., Ruiz-Fernández, J., Barriendos, M., Benito, G., Cuadrat, J.M., Domínguez-Castro, F., 1357 
García-Ruiz, J.M., Giralt, S., Gómez-Ortiz, A., Hernández, A., López-Costas, O., López-Moreno, 1358 
J.I., López-Sáez, J.A., Martínez-Cortizas, A., Moreno, A., Prohom, M., Saz, M.A., Serrano, E., 1359 
Tejedor, E., Trigo, R., Valero-Garcés, B., Vicente-Serrano, S.M., 2018. The Little Ice Age in 1360 
Iberian mountains. Earth-Sci. Rev. 177, 175–208. 1361 
https://doi.org/10.1016/j.earscirev.2017.11.010. 1362 

Orwin, J.F., Mckinzey, K.M., Stephens, M.A., Dugmore, A.J., 2008. Identifying moraine surfaces 1363 
with similar histories using lichen size distributions and the U2 statistic, Southeast Iceland. Geogr. 1364 
Ann. Ser. A Phys. Geogr. 90 (A), 151–164. https://doi.org/10.1111/j.1468-0459.2008.00168.x. 1365 

Osborn, G., McCarthy, D., LaBrie, A., Burke, R., 2015. Lichenometric dating: science or pseudo-1366 
science? Quat. Res. 83, 1–12. https://doi.org/10.1016/j.yqres.2014.09.006. 1367 

Osmaston, H., 2005. Estimates of glacier equilibrium line altitudes by the Area × Altitude, the 1368 
Area × Altitude Balance Ratio and the Area × Altitude Balance Index methods and their 1369 
validation. Quat. Int. 138–139, 22–31. https://doi.org/10.1016/j.quaint.2005.02.004. 1370 

Palacios, D., Gómez-Ortiz, A., Alcalá-Reygosa, J., Andrés, N., Oliva, M., Tanarro, L.M., 1371 
Salvador-Franch, F., Schimmelpfennig, I., Fernández-Fernández, J.M., Léanni, L., 2019. The 1372 
challenging application of cosmogenic dating methods in residual glacial landforms: the case of 1373 

https://doi.org/10.1177/0959683612463104
https://doi.org/10.1029/2011GL050168
https://www.lmi.is/landupplysingar/loftmyndasafn-2-2/
https://doi.org/10.14430/arctic869
https://doi.org/10.1016/j.earscirev.2017.11.010
https://doi.org/10.1111/j.1468-0459.2008.00168.x
https://doi.org/10.1016/j.yqres.2014.09.006
https://doi.org/10.1016/j.quaint.2005.02.004


Sierra Nevada (Spain). Geomorphology 325, 103–118. 1374 
https://doi.org/10.1016/j.geomorph.2018.10.006. 1375 

Paterson, W.S.B., 1994. The Physics of Glaciers. 3rd edition. Pergamon/Elsevier, London. 1376 

Pellitero, R., Rea, B.R., Spagnolo, M., Bakke, J., Hughes, P., Ivy-Ochs, S., Lukas, S., Ribolini, 1377 
A., 2015. A GIS tool for automatic calculation of glacier equilibrium-line altitudes. Comput. 1378 
Geosci. 82, 55–62. https://doi.org/10.1016/j.cageo.2015.05.005. 1379 

Pellitero, R., Rea, B.R., Spagnolo, M., Bakke, J., Ivy-Ochs, S., Frew, C.R., Hughes, P., Ribolini, 1380 
A., Lukas, S., Renssen, H., 2016. GlaRe, a GIS tool to reconstruct the 3D surface of 1381 
palaeoglaciers. Comput. Geosci. 94, 77–85. https://doi.org/10.1016/j.cageo.2016.06.008. 1382 

Principato, S.M., Geirsdóttir, Á., Jóhannsdóttir, G.E., Andrews, J.T., 2006. Late Quaternary 1383 
glacial and deglacial history of eastern Vestfirðir, Iceland using cosmogenic isotope (36Cl) 1384 
exposure ages and marine cores. J. Quat. Sci. 21, 271–285. https://doi.org/10.1002/jqs.978. 1385 

Rea, B.R., 2009. Defining modern day Area-Altitude Balance Ratios (AABRs) and their use in 1386 
glacier-climate reconstructions. Quat. Sci. Rev. 28, 237–248. 1387 
https://doi.org/10.1016/j.quascirev.2008.10.011. 1388 

Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Buck, C.E., 1389 
Cheng, H., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Haflidason, H., Hajdas, 1390 
I., Hatté, C., Heaton, T.J., Hoffmann, D.L., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., 1391 
Manning, S.W., Niu, M., Reimer, R.W., Richards, D.A., Scott, E.M., Southon, J.R., Staff, R.A., 1392 
Turney, C.S.M., van der Plicht, J., 2013. IntCal13 and Marine13 radiocarbon age calibration 1393 
curves 0–50,000 years cal BP. Radiocarbon 55 (4), 1869–1887. 1394 

Roca-Valiente, B., Hawksworth, D.L., Pérez-Ortega, S., Sancho, L.G., Crespo, A., 2016. Type 1395 
studies in the Rhizocarpon geographicum group (Rhizocarpaceae, lichenized Ascomycota). 1396 
Lichenologist 48, 97–110. https://doi.org/10.1017/S002428291500050X. 1397 

Rogers, J.C., Van Loon, H., 1979. The seesaw in winter temperatures between Greenland and 1398 
Northern Europe. Part II: some oceanic and atmospheric effects in middle and high latitudes. 1399 
Mon. Weather Rev. 107, 509–519. https://doi.org/10.1175/1520-1400 
0493(1979)107b0509:TSIWTBN2.0.CO;2. 1401 

Russell, A.J., Knight, P.G., Van Dijk, T.A.G.P., 2001. Glacier surging as a control on the 1402 
development of proglacial, fluvial landforms and deposits, Skeiðarársandur, Iceland. Glob. 1403 
Planet. Chang. 28, 163–174. https://doi.org/10.1016/S0921-8181(00)00071-0. 1404 

Sæmundsson, K., Kristjansson, L., McDougall, I., Watkins, N.D., 1980. K-Ar dating, geological 1405 
and paleomagnetic study of a 5-km lava succession in northern Iceland. J. Geophys. Res. Solid 1406 
Earth 85, 3628–3646. https://doi.org/10.1029/JB085iB07p03628. 1407 

Sæmundsson, Þ., Morino, C., Helgason, J.K., Conway, S.J., Pétursson, H.G., 2018. The triggering 1408 
factors of the Móafellshyrna debris slide in northern Iceland: intense precipitation, earthquake 1409 
activity and thawing of mountain permafrost. Sci. Total Environ. 621, 1163–1175. 1410 
https://doi.org/10.1016/J.SCITOTENV.2017.10.111. 1411 

Sancho, L.G., Palacios, D., Green, T.G.A., Vivas, M., Pintado, A., 2011. Extreme high lichen 1412 
growth rates detected in recently deglaciated areas in Tierra del Fuego. Polar Biol. 34, 813–822. 1413 
https://doi.org/10.1007/s00300-010-0935-4. 1414 

https://doi.org/10.1016/j.geomorph.2018.10.006
https://doi.org/10.1016/j.cageo.2015.05.005
https://doi.org/10.1016/j.cageo.2016.06.008
https://doi.org/10.1002/jqs.978
https://doi.org/10.1016/j.quascirev.2008.10.011
https://doi.org/10.1017/S002428291500050X
https://doi.org/10.1175/1520-0493(1979)107b0509:TSIWTBN2.0.CO;2
https://doi.org/10.1175/1520-0493(1979)107b0509:TSIWTBN2.0.CO;2
https://doi.org/10.1016/S0921-8181(00)00071-0
https://doi.org/10.1029/JB085iB07p03628
https://doi.org/10.1016/J.SCITOTENV.2017.10.111
https://doi.org/10.1007/s00300-010-0935-4


Sancho, L.G., Pintado, A., Navarro, F., Ramos, M., De Pablo, M.A., Blanquer, J.M., Raggio, J., 1415 
Valladares, F., Green, T.G.A., 2017. Recent warming and cooling in the Antarctic Peninsula 1416 
Region has rapid and large effects on lichen vegetation. Sci. Rep. 7, 5689. 1417 
https://doi.org/10.1038/s41598-017-05989-4. 1418 

Schaefer, J.M., Denton, G.H., Kaplan, M., Putnam, A., Finkel, R.C., Barrell, D.J.A., Andersen, 1419 
B.G., Schwartz, R., Mackintosh, A., Chinn, T., Schlüchter, C., 2009. High-frequency Holocene 1420 
glacier fluctuations in New Zealand differ from the northern signature. Science 324, 622–625. 1421 
https://doi.org/10.1126/science.1169312. 1422 

Schildgen, T.F., Phillips, W.M., Purves, R.S., 2005. Simulation of snow shielding corrections for 1423 
cosmogenic nuclide surface exposure studies. Geomorphology 64, 67–85. 1424 
https://doi.org/10.1016/j.geomorph.2004.05.003. 1425 

Schimmelpfennig, I., Benedetti, L., Finkel, R., Pik, R., Blard, P.H., Bourlès, D., Burnard, P., 1426 
Williams, A., 2009. Sources of in-situ 36Cl in basaltic rocks. Implications for calibration of 1427 
production rates. Quat. Geochronol. 4, 441–461. https://doi.org/10.1016/j.quageo.2009.06.003. 1428 

Schimmelpfennig, I., Benedetti, L., Garreta, V., Pik, R., Blard, P.H., Burnard, P., Bourlès, D., 1429 
Finkel, R., Ammon, K., Dunai, T., 2011. Calibration of cosmogenic 36Cl production rates from 1430 
Ca and K spallation in lava flows from Mt. Etna (38°N, Italy) and Payun Matru (36°S, Argentina). 1431 
Geochim. Cosmochim. Acta 75, 2611–2632. https://doi.org/10.1016/j.gca.2011.02.013. 1432 

Schimmelpfennig, I., Schaefer, J.M., Akçar, N., Ivy-Ochs, S., Finkel, R.C., Schlüchter, C., 2012. 1433 
Holocene glacier culminations in the Western Alps and their hemispheric relevance. Geology 40, 1434 
891–894. https://doi.org/10.1130/G33169.1. 1435 

Schimmelpfennig, I., Schaefer, J.M., Akçar, N., Koffman, T., Ivy-Ochs, S., Schwartz, R., Finkel, 1436 
R.C., Zimmerman, S., Schlüchter, C., 2014a. A chronology of Holocene and Little Ice Age glacier 1437 
culminations of the Steingletscher, Central Alps, Switzerland, based on high-sensitivity 1438 
beryllium-10 moraine dating. Earth Planet. Sci. Lett. 393, 220–230. 1439 
https://doi.org/10.1016/j.epsl.2014.02.046. 1440 

Schimmelpfennig, I., Schaefer, J.M., Putnam, A.E., Koffman, T., Benedetti, L., Ivy-Ochs, S., 1441 
Team, A., Schlüchter, C., 2014b. 36 Cl production rate from K-spallation in the European Alps 1442 
(Chironico landslide, Switzerland). J. Quat. Sci. 29, 407–413. https://doi.org/10.1002/jqs.2720. 1443 

Schomacker, A., Krüger, J., Larsen, G., 2003. An extensive late Holocene glacier advance of 1444 
Kötlujökull, central south Iceland. Quat. Sci. Rev. 22, 1427–1434. https://doi.org/10.1016/S0277-1445 
3791(03)00090-8. 1446 

Schomacker, A., Brynjólfsson, S., Andreassen, J.M., Gudmundsdóttir, E.R., Olsen, J., Odgaard, 1447 
B.V., Håkansson, L., Ingólfsson, O., Larsen, N.K., 2016. The Drangajökull ice cap, northwest 1448 
Iceland, persisted into the early-mid Holocene. Quat. Sci. Rev. 148, 66–84. 1449 

Sigurðsson, O., 1998. Glacier variations in Iceland 1930–1995. Jökull 45, 27–33. 1450 

Sigurðsson, O., 2005. 10. Variations of termini of glaciers in Iceland in recent centuries and their 1451 
connection with climate. Developments in Quaternary Science, pp. 241–255 1452 
https://doi.org/10.1016/S1571-0866(05)80012-0. 1453 

https://doi.org/10.1038/s41598-017-05989-4
https://doi.org/10.1126/science.1169312
https://doi.org/10.1016/j.geomorph.2004.05.003
https://doi.org/10.1016/j.quageo.2009.06.003
https://doi.org/10.1016/j.gca.2011.02.013
https://doi.org/10.1130/G33169.1
https://doi.org/10.1016/j.epsl.2014.02.046
https://doi.org/10.1002/jqs.2720
https://doi.org/10.1016/S0277-3791(03)00090-8
https://doi.org/10.1016/S0277-3791(03)00090-8
https://doi.org/10.1016/S1571-0866(05)80012-0


Sigurðsson, O., Jónsson, T., Jóhannesson, T., 2007. Relation between glacier-termini variations 1454 
and summer temperature in Iceland since 1930. Ann. Glaciol. 46, 170–176. 1455 
https://doi.org/10.3189/172756407782871611. 1456 

Sissons, J.B., 1974. A Late-glacial ice cap in the Central Grampians, Scotland. Trans. Inst. Br. 1457 
Geogr. 95. https://doi.org/10.2307/621517. 1458 

Solomina, O.N., Bradley, R.S., Jomelli, V., Geirsdottir, A., Kaufman, D.S., Koch, J.,McKay, N.P., 1459 
Masiokas, M., Miller, G., Nesje, A., Nicolussi, K., Owen, L.A., Putnam, A.E., Wanner, H., Wiles, 1460 
G., Yang, B., 2016. Glacier fluctuations during the past 2000 years. Quat. Sci. Rev. 149, 61–90. 1461 
https://doi.org/10.1016/j.quascirev.2016.04.008. 1462 

Stone, J.O., 2000. Air pressure and cosmogenic isotope production. J. Geophys. Res. Solid Earth 1463 
105, 23753–23759. https://doi.org/10.1029/2000JB900181. 1464 

Stone, J.O., Allan, G.L., Fifield, L.K., Cresswell, R.G., 1996. Cosmogenic chlorine-36 from 1465 
calcium spallation. Geochim. Cosmochim. Acta 60, 679–692. https://doi.org/10.1016/0016-1466 
7037(95)00429-7. 1467 

Stone, J.O., Fifielde, K., Vasconcelos, P., 2005. Terrestrial chlorine-36 production from spallation 1468 
of iron. Abstract of 10th International Conference on Accelerator Mass Spectrometry (Berkeley, 1469 
CA). 1470 

Stötter, J., 1990. Geomorphologische und landschaftsgeschichtliche Untersuchungen im 1471 
Svarfaðardalur-Skíðadalur, Tröllaskagi, N-Island. Munch. Geogr. Abh. 9. 1472 

Stötter, J., 1991. New observations on the postglacial glacial history of Tröllaskagi, Northern 1473 
Iceland. In: Maizels, J.K., Caseldine, C. (Eds.), Environmental Change in Iceland: Past and 1474 
Present. Glaciology and Quaternary Geology vol. 7. Springer, Dordrecht, pp. 181–192. 1475 
https://doi.org/10.1007/978-94-011-3150-6_12. 1476 

Stötter, J., Wastl, M., Caseldine, C., Häberle, T., 1999. Holocene palaeoclimatic reconstruction in 1477 
northern Iceland: approaches and results. Quat. Sci. Rev. 18, 457–474. 1478 
https://doi.org/10.1016/S0277-3791(98)00029-8. 1479 

Striberger, J., Björck, S., Holmgren, S., Hamerlík, L., 2012. The sediments of Lake Lögurinn - A 1480 
unique record of Holocene glacial meltwater variability in eastern Iceland. Quat. Sci. Rev. 38, 1481 
76–88. https://doi.org/10.1016/j.quascirev.2012.02.001. 1482 

Tanarro, L.M., Palacios, D., Andrés, N., Fernández-Fernández, J.M., Zamorano, J.J., 1483 
Sæmundsson, Þ., Brynjólfsson, S., 2019. Unchanged surface morphology in debris-covered 1484 
glaciers and rock glaciers in Tröllaskagi peninsula (northern Iceland). Sci. Total Environ. 648, 1485 
218–235. https://doi.org/10.1016/j.scitotenv.2018.07.460. 1486 

Thompson, A., Jones, A., 1986. Rates and causes of proglacial river terrace formation in southeast 1487 
Iceland: an application of lichenometric dating techniques. Boreas 15, 231–246. 1488 
https://doi.org/10.1111/j.1502-3885.1986.tb00928.x. 1489 

Van der Veen, C.J., 1999. Fundamentals of Glacier Dynamics. Balkema, Rotterdam. 1490 

Wagner, S., Zorita, E., 2005. The influence of volcanic, solar and CO2 forcing on the temperatures 1491 
in the Dalton Minimum (1790–1830): a model study. Clim. Dyn. 25, 205–218. 1492 
https://doi.org/10.1007/s00382-005-0029-0. 1493 

https://doi.org/10.3189/172756407782871611
https://doi.org/10.2307/621517
https://doi.org/10.1016/j.quascirev.2016.04.008
https://doi.org/10.1029/2000JB900181
https://doi.org/10.1016/0016-7037(95)00429-7
https://doi.org/10.1016/0016-7037(95)00429-7
https://doi.org/10.1007/978-94-011-3150-6_12
https://doi.org/10.1016/S0277-3791(98)00029-8
https://doi.org/10.1016/j.quascirev.2012.02.001
https://doi.org/10.1016/j.scitotenv.2018.07.460
https://doi.org/10.1111/j.1502-3885.1986.tb00928.x
https://doi.org/10.1007/s00382-005-0029-0


Wanner, H., Beer, J., Bütikofer, J., Crowley, T.J., Cubasch, U., Flückiger, J., Goosse, H., Grosjean, 1494 
M., Joos, F., Kaplan, J.O., Küttel, M., Müller, S.A., Prentice, I.C., Solomina, O., Stocker, T.F., 1495 
Tarasov, P., Wagner, M., Widmann, M., 2008. Mid- to Late Holocene climate change: an overview. 1496 
Quat. Sci. Rev. 27, 1791–1828. https://doi.org/10.1016/j.quascirev.2008.06.013. 1497 

Wastl, M., Stötter, J., 2005. 9. Holocene glacier history. Developments in Quaternary Science, pp. 1498 
221–240 https://doi.org/10.1016/S1571-0866(05)80011-9. 1499 

Whalley, W.B., Douglas, G.R., Jonsson, A., 1983. The magnitude and frequency of large 1500 
rockslides in Iceland in the postglacial. Geogr. Ann. Ser. A., Phys. Geogr. 65, 99–110. 1501 
https://doi.org/10.2307/520724. 1502 

Wiles, G.C., Barclay, D.J., Young, N.E., 2010. A review of lichenometric dating of glacial 1503 
moraines in Alaska. Geogr. Ann. Ser. A Phys. Geogr. 92, 101–109. https://doi.org/10.1111/j.1468-1504 
0459.2010.00380.x. 1505 

Winkler, S., 2003. A new interpretation of the date of the “Little Ice Age” glacier maximum at 1506 
Svartisen and Okstindan, northern Norway. The Holocene 13, 83–95. 1507 
https://doi.org/10.1191/0959683603hl573rp. 1508 

Xiao, X., Zhao,M., Knudsen, K.L., Sha, L., Eiríksson, J., Gudmundsdóttir, E., Jiang, H., Guo, Z., 1509 
2017. Deglacial and Holocene sea–ice variability north of Iceland and response to ocean 1510 
circulation changes. Earth Planet. Sci. Lett. 472, 14–24. 1511 
https://doi.org/10.1016/J.EPSL.2017.05.006. 1512 

Young, N.E., Schweinsberg, A.D., Briner, J.P., Schaefer, J.M., 2015. Glacier maxima in Baffin 1513 
Bay during the Medieval Warm Period coeval with Norse settlement. Sci. Adv. 1. 1514 
https://doi.org/10.1126/sciadv.1500806. 1515 

Zhong, Y., Miller, G.H., Otto-Bliesner, B.L., Holland, M.M., Bailey, D.A., Schneider, D.P., 1516 
Geirsdóttir, A., 2011. Centennial-scale climate change from decadally-paced explosive 1517 
volcanism: a coupled sea ice-ocean mechanism. Clim. Dyn. 37, 2373–2387. 1518 
https://doi.org/10.1007/s00382-010-0967-z. 1519 

Zweck, C., Zreda,M., Desilets, D., 2013. Snow shielding factors for cosmogenic nuclide dating 1520 
inferred from Monte Carlo neutron transport simulations. Earth Planet. Sci. Lett. 379, 64–71. 1521 
https://doi.org/10.1016/J.EPSL.2013.07.023. 1522 

 1523 

https://doi.org/10.1016/j.quascirev.2008.06.013
https://doi.org/10.1016/S1571-0866(05)80011-9
https://doi.org/10.2307/520724
https://doi.org/10.1111/j.1468-0459.2010.00380.x
https://doi.org/10.1111/j.1468-0459.2010.00380.x
https://doi.org/10.1191/0959683603hl573rp
https://doi.org/10.1016/J.EPSL.2017.05.006
https://doi.org/10.1126/sciadv.1500806
https://doi.org/10.1007/s00382-010-0967-z
https://doi.org/10.1016/J.EPSL.2013.07.023


Table 1. Glacial Equilibrium-Line Altitudes (ELAs) calculated over Tungnahryggsjökull 1524 
glaciers through the application of the AAR and AABR methods over the 3D glacier 1525 
reconstructions. Delta (Δ) is referred to the change with respect to the previous stage. 1526 

Stage   W Tungnahryggsjökull   E Tungnahryggsjökull 

  AAR (0.67) Δ AABR (1.5±0.4) Δ   AAR (0.67) Δ AABR (1.5±0.4) Δ 

1   1021 - 1006 +25/-20 -   1032 - 1027 ±20 - 

2   1047 +26 1032 ±20 +26   1032 0 1032 ±15 +5 

3   1052 +5 1037 +20/-15 +5   1034 +2 1034 ±15 +2 

4   1054 +2 1049 +20/-15 +12   1037 +3 1037 ±15 +3 

5   1059 +5 1059 ±15 +10   1041 +4 1041 ±15 +4 

6   1067 +8 1062 +20/-10 +3   1042 +1 1042 ±15 +1 

7   1072 +5 1072 +15/-10 +10   1046 +5 1046 ±15 +5 

8   1076 +4 1081 ±15 +9   1047 +1 1052 ±15 +6 

9   1082 +6 1087 +15/-10 +6   1054 +7 1054 +15/-10 +2 

10   1086 +4 1091 +15/-10 +4   1055 +1 1060 +15/-10 +6 

11   1094 +8 1099 +15/-10 +8   1061 +6 1071 ±10 +11 

12   1094 0 1099 +15/-5 0   1060 -1 1070 ±10 -1 

13   1094 0 1099 +15/-5 0   1061 +1 1071 +15/-10 +1 

14   1100 +6 1110 ±10 +11   1060 -1 1075 ±10 +4 

15   1102 +2 1107 +15/-5 -3   1061 +1 1076 +15/-10 +1 

16   1098 -4 1108 +10/-5 +1   1061 0 1076 ±10 0 

17   1099 +1 1109 +15/-5 +1   1065 +4 1080 ±10 +4 

 1527 



Table 2. Surface ages estimated from Kugelmann’s (1991) 0.44 mm yr−1 growth rate for 1528 
different colonization lags. The dates obtained from Rhizocarpon geographicum lichens 1529 
discussed throughout the text are those derived from a 20-yr colonization lag. The figures in 1530 
italics correspond to ages tentatively inferred from the largest Porpidia soredizodes lichen, 1531 
assuming a 0.737 mm yr−1 growth rate and a 15-year colonization lag (see section 5.2.1). 1532 

Glacier foreland Lichen 

station 

Glacial 

stage 

Date 

deduced 

from 

photographic 

evidence 

Surface age from growth rate (yr) 

10-yr 

col. lag 

15-yr 

col. lag. 

20-yr 

col. lag 

25-yr 

col. lag. 

30-yr 

col. lag 

W 

Tungnahryggsjökull 
TUW-1 15-16 1994-2000 35a 40a 45a 50a 55a 

TUW-2 13-15 1946-1994 54 59 64 69 74 a 

TUW-3 12-13 <1946 65 a 70 75 80 85 

TUW-4 12 <1946 90 95 100 105 110 

TUW-5 11 <1946 111 116 121 126 131 

TUW-6 10 <1946 161 166 171 176 181 

TUW-7 9 <1946 173 178 183 188 193 

TUW-8 8 <1946 211 216 221 226 231 

E 

Tungnahryggsjökull 
TUE-0 post-17 2005< - - - - - 

TUE-1 14 1946-1985 33 38 43 48 53 

TUE-2 13 1946-1985 53 58 63 68 73 a 

TUE-3 12 1946-1985 72 a 77 a 82 87 a 92 a 

TUE-4 11 <1946 90 95 100 105 110 

TUE-5 10 <1946 136 141 146 151 156 

TUE-6 9 <1946 114 b 119 b 124 b 129 b 134 b 

TUE-7 7 <1946 119 b 124 b 129 b 134 b 139 b 

TUE-8 7 <1946 202 207 212 217 222 

 1533 

a The age does not agree with the date deduced from aerial photos. 1534 

b The age is incoherent with the moraine chronostratigraphy. 1535 

 1536 



Table 3. 36Cl CRE ages converted to CE/BCE dates according to the different 36Cl 1537 
production rates from Ca spallation. Uncertainties include the analytical and production 1538 
rate error. 1539 

Sample name Dates (CE/BCE) 

Licciardi et al. (2008) Ca 

spallation prod. rate 

Stone et al. (1996) Ca 

spallation prod. rate 

Schimmelpfennig et al. 

(2011) Ca spallation 

prod. rate 

Moraine boulders at Vesturdalur (W Tungnahryggsjökull foreland) 

TUW-9 1590 ± 100 1570±100 1520±120 

TUW-10 1640±90 1620±90 1580±110 

TUW-11 1480±120 1470±120 1420±140 

TUW-12 1450±100 1430±110 1370±120 

TUW-13 1470±130 1450±130 1400±150 

TUW-14 670±210 620±210 500±250 

TUW-15 1460±110 1430±110 1370±120 

TUW-16 1220±190 1200±190 1130±210 

Moraine boulders at Austurdalur (E Tungnahryggsjökull foreland) 

TUE-9 740±170 660±170 510±210 

TUE-10 1460±100 1430±100 1360±120 

TUE-11 380±200 290±200 110±250 

TUE-12 400±200 320±190 150±240 

Glacially polished ridge Elliði 

ELLID-1 14300±1700 14900±1600 16600±2100 

ELLID-2 14200±1700 14800±1700 16500±2100 

 1540 



Fig. 1. Location of the Tungnahryggsjökull glaciers and their forelands (C) (Vesturdalur and 1541 
Austurdalur) in the context of Iceland (A) and the Tröllaskagi peninsula (B). The red boxes 1542 
in panels A and B are panels B and C, respectively. The figure also includes the location of 1543 
the place names mentioned throughout the paper. This figure is available in colour in the 1544 
online version. 1545 

 1546 



Fig. 2. Geomorphological map of the Vesturdalur foreland. (A) General view of the Western 1547 
Tungnahryggsjökull foreland. (B) Detailed moraine mapping and glacier margin geometry 1548 
reconstructed throughout the different glacial stages identified, and 36Cl CRE and 1549 
lichenometric dating results (both are expressed in ages and calendar years). Note that 1550 
stages 13, 14, 15 and 16 correspond to the years 1946, 1985, 1994 and 2000. The surface-1551 
contoured glacier (white) corresponds to the year 2005 (stage 17). The abbreviations “Rg” 1552 
and “Ps” in lichenometry stations indicate that the estimated dates are derived from the 1553 
Rhizocarpon geographicum and Porpidia soredizodes lichens, respectively, and the number 1554 
correspond to the longest axis of the largest lichen measured. 1555 

 1556 



Fig. 3. Geomorphological map of the Austurdalur foreland. (A) General view of the Eastern 1557 
Tungnahryggsjökull foreland. (B) Detailed moraine mapping and glacier margin geometry 1558 
reconstructed throughout the different glacial stages identified, and 36Cl CRE and 1559 
lichenometric dating results (both are expressed in ages and calendar years). Note that 1560 
stages 11, 15 and 16 correspond to the years 1946, 1985 and 2000. The surface-contoured 1561 
glacier (white) corresponds to the year 2005 (stage 17). The abbreviations “Rg” and “Ps” in 1562 
lichenometry stations indicate that the estimated dates are derived from the Rhizocarpon 1563 
geographicum and Porpidia soredizodes lichens, respectively, and the number correspond to 1564 
the longest axis of the largest lichen measured. 1565 

 1566 



Fig. 4. 36Cl CRE ages and internal (analytical) uncertainty at 1σ level of the samples from 1567 
Vesturdalur and Austurdalur. Note that the samples clustering around 500 yr (15th century) 1568 
in Austurdalur are indistinguishable. Distance to the terminus (year 2005) is measured along 1569 
the flowline from the reconstructed snout apex of the phase where the samples were 1570 
collected. This figure is available in colour in the online version. 1571 

 1572 



SUPPLEMENTARY TABLES 1573 



ST1. Correspondence between glacial stages mapped over historical aerial photos and the 1574 
dates. 1575 

W Tungnahryggsjökull   E Tungnahryggsjökull 

Stage Date   Stage Date 

13 1946   11 1946 

14 1985   15 1985 

15 1994   - - 

16 2000   16 2000 

17 2005   17 2005 

 1576 



ST2. Glacier length and snout position variations measured along the flowline. The asterisk 1577 
(*) indicates uneven changes in the snout, with sectors retreating or advances. Delta (Δ) is 1578 
referred to the change with respect to the previous stage. 1579 

Stage   W Tungnahryggsjökull   E Tungnahryggsjökull 

  Length (m) Δ (m)   Length (m) Δ (m) 

1   6492 -   3777 - 

2   5483 -1009   3733 -44 

3   5345 -138   3627 -106 

4   4946 -398   3515 -112 

5   4761 -185   3398 -117 

6   4682 -79   3362 -36 

7   4402 -280   3237 -125 

8   4218 -185   3084 -153 

9   3951 -266   3057 -27 

10   3800 -151   2911 -146 

11   3601 -199   2755 -156 

12   3452 -149   2722 -34 

13   3406 -46   2618 -103 

14   3338 -69   2531 -87 

15   3311 -27 (*)   2477 -54 

16   3203 -108   2513 +36 

17   3183 -20   2485 -29 

 1580 



ST3. Glacier extent and volume variations calculated from 2D and 3D glacier 1581 
reconstructions. Delta (Δ) refers to the change with respect to the previous stage. 1582 

Stage   W Tungnahryggsjökull   E Tungnahryggsjökull 

  

Area 

(km2) 

Δ (%) Vol. 

(km3) 

Δ (%)   Area 

(km2) 

Δ (%) Vol. 

(km3) 

Δ (%) 

1   9.44 - 1.103 -   5.28 - 0.466 - 

2   8.87 -6.04 0.998 -9.55   5.26 -0.29 0.463 -0.49 

3   8.76 -1.22 0.985 -1.33   5.22 -0.82 0.454 -1.91 

4   8.46 -3.45 0.942 -4.34   5.15 -1.26 0.446 -1.97 

5   8.36 -1.18 0.928 -1.47   5.09 -1.19 0.434 -2.52 

6   8.14 -2.65 0.918 -1.05   5.06 -0.52 0.433 -0.27 

7   7.80 -4.14 0.889 -3.24   4.98 -2.26 0.422 -2.84 

8   7.64 -2.05 0.870 -2.15   4.88 -1.97 0.409 -2.99 

9   7.32 -4.13 0.842 -3.22   4.81 -1.40 0.407 -0.66 

10   7.18 -2.02 0.826 -1.86   4.71 -2.13 0.395 -2.77 

11   6.98 -2.68 0.813 -1.58   4.56 -3.17 0.387 -2.19 

12   6.94 -0.67 0.804 -1.06   4.54 -0.42 0.383 -0.86 

13   6.91 -0.37 0.801 -0.49   4.51 -0.73 0.381 -0.67 

14   6.67 -3.47 0.790 -1.27   4.05 -10.12 0.363 -4.78 

15   6.68 +0.17 0.791 +0.09   4.02 -0.65 0.361 -0.46 

16   6.57 -1.63 0.774 -2.11   4.02 -0.14 0.362 +0.16 

17   6.47 -1.57 0.771 -0.43   3.88 -3.45 0.360 -0.52 

 1583 



ST4. Size of the largest lichen at the lichenometry stations in the forelands of the western 1584 
and eastern Tungnahryggsjökull glaciers. 1585 

Glacier foreland   Lichen 

station 

  GPS location  Glacial 

stage 
  

Rhizocarpon 

geographicum 

 
Porpidia 

soredizodes 

    

Latitude 

(N)  

Longitude 

(W) 

   Min. circ. 

diameter (mm) 

 
Min. circ. 

diameter 

(mm) 

W 

Tungnahryggsjökull 

(Vesturdalur) 

  TUW-1   65º41.874’ 18º52.785’  15-16   - a   18.3 

  TUW-2   65º41.899’ 18º52.749’  13-15   19.3   43.7 

  TUW-3   65º41.968’ 18º52.686’  12-13   24.3   42.8 

  TUW-4   65º41.979’ 18º52.673’  12   35.2   62.1 

  TUW-5   65º42.045’ 18º52.571’  11   44.5   69.6 

  TUW-6   65º42.097’ 18º52.539’  10   66.6   106.4 

  TUW-7   65º42.197’ 18º52.451’  9   71.5   106.2 

  TUW-8   65º42.313’ 18º52.411’  8   -  148.4 

E 

Tungnahryggsjökull 

(Austurdalur) 

 TUE-0  65º42.109’ 18º49.383’  post-17  -  - 

  TUE-1   65º42.125’ 18º49.448’  14   -   16.8 

  TUE-2   65º42.145’ 18º49.502’  13   19.0   39.4 

  TUE-3   65º42.186’ 18º49.590’  12   27.1   53.2 

  TUE-4   65º42.200’ 18º49.628’  11   35.2   60.8 

  TUE-5   65º42.248’ 18º49.697’  10   55.6   128.2 

  TUE-6   65º42.331’ 18º49.760’  9   45.7   116.9 

  TUE-7   65º42.446’ 18º49.883’  7   47.7   100.8 

  TUE-8   65º42.495’ 18º49.810’  7   -   141.8 

 1586 



ST5. Geographic sample locations, topographic shielding factor, sample thickness and 1587 
distance from terminus. 1588 

Sample 

name 

Sta

ge 

Latitude 

(ºN) 

Longitude 

(ºW) 

Elevation (m 

a.s.l.) 

Shielding 

factor 

Thickness 

(cm) 

Dist. to 

terminus (m) a 

Moraine boulders at Vesturdalur (W Tungnahryggsjökull foreland) 

TUW-9 8 65.7053 18.8744 674 0.9662 4.0 1034 

TUW-10 8 65.7055 18.8740 669 0.9818 4.0 1034 

TUW-11 7 65.7093 18.8725 599 0.9805 2.0 1219 

TUW-12 6 65.7098 18.8728 593 0.9799 3.5 1499 

TUW-13 4 65.7156 18.8725 522 0.9683 3.0 1763 

TUW-14 4 65.7156 18.8725 523 0.9674 2.0 1763 

TUW-15 3 65.7168 18.8734 523 0.9788 3.0 2161 

TUW-16 3 65.7169 18.8733 520 0.9798 2.0 2161 

Moraine boulders at Austurdalur (E Tungnahryggsjökull foreland)  

TUE-9 4 65.7103 18.8321 624 0.9848 3.0 1030 

TUE-10 4 65.7104 18.8321 623 0.9849 2.5 1030 

TUE-11 1 65.7101 18.8364 692 0.9667 3.0 1293 

TUE-12 1 65.7100 18.8364 694 0.9756 3.5 1293 

Glacially polished ridge separating Viðinesdalur and Kolbeinsdalur     

ELLID-1 - 65.75800 19.0848 608 0.9989 2.0 14370 

ELLID-2 - 65.75790 19.0854 611 0.9993 3.5 14400 

 1589 

a Distance to the terminus (year 2005) is measured along the flowline from the reconstructed snout 1590 
apex of the phase where the samples were collected. 1591 



ST6. Chemical composition of the bulk rock samples before chemical treatment. The figures in italics correspond to the average values of the bulk 1592 
samples analysed and were those used for the age-exposure calculations of those samples without bulk chemical composition analysis. 1593 

Sample 

name 

CaO 

(%) 

K2O 

(%) 

TiO2 

(%) 

Fe2O3 

(%) 

Cl 

(ppm) 

SiO2 

(%) 

Na2O 

(%) 

MgO 

(%) 

Al2O3 

(%) 

MnO 

(%) 

P2O5 

(%) 

CO2 

(%) 

Li 

(ppm) 

B 

(ppm

) 

Sm 

(ppm) 

Gd 

(ppm) 

Th 

(ppm) 

U 

(ppm) 

Moraine boulders at Vesturdalur (W Tungnahryggsjökull foreland)  

TUW-15 10.540 0.300 3.704 15.510 27 48.800 2.330 5.966 11.913 0.217 < L.D. - 4.8 <2 2.379 2.892 0.622 0.225 

TUW-16 10.388 0.377 3.239 14.930 44 49.690 2.113 6.937 11.433 0.211 < L.D. - 3.9 <2 2.121 2.523 0.910 0.306 

Average 10.464 0.339 3.472 15.220 36 49.245 2.222 6.452 11.673 0.214 < L.D. - 4.4 <2 2.250 2.707 0.766 0.265 

Moraine boulders at Austurdalur (E Tungnahryggsjökull foreland)  

TUE-11 12.333 0.199 2.233 12.720 43 47.270 2.106 7.268 14.465 0.183 0.180 - 3.6 <2 4.204 4.402 0.821 0.224 

TUE-12 12.325 0.229 2.219 12.617 43 47.680 2.284 6.988 14.710 0.182 0.190 - 3.6 <2 4.392 4.525 0.864 0.216 

Average 12.329 0.214 2.226 12.669 43 47.475 2.195 7.128 14.588 0.182 0.185 - 3.6 <2 4.298 4.464 0.842 0.220 

Glacially polished ridge separating Viðinesdalur and Kolbeinsdalur 

ELLID-1 8.568 0.179 4.753 18.215 21 44.950 2.161 7.475 12.416 0.228 0.200 - 4.4 2 5.709 5.953 0.504 0.149 

 1594 



ST7. Concentrations of the 36Cl target elements Ca, K, Ti and Fe, determined in splits taken 1595 
after the chemical pre-treatment (acid etching). 1596 

Sample name CaO (%) K2O (%) TiO2 (%) Fe2O3 (%) 

Moraine boulders at Vesturdalur (W Tungnahryggsjökull foreland) 

TUW-9 10.96 ± 0.22 0.27 ± 0.04 3.57 ± 0.18 13.46 ± 0.27 

TUW-10 11.06 ± 0.22 0.25 ± 0.04 3.07 ± 0.15 14.31 ± 0.29 

TUW-11 10.57 ± 0.21 0.37 ± 0.06 3.72 ± 0.19 15.32 ± 0.31 

TUW-12 11.78 ± 0.24 0.32 ± 0.05 2.92 ± 0.15 13.62 ± 0.27 

TUW-13 10.51 ± 0.21 0.37 ± 0.06 3.66 ± 0.18 15.27 ± 0.31 

TUW-14 9.98 ± 0.20 0.36 ± 0.05 3.43 ± 0.17 16.16 ± 0.32 

TUW-15 10.75 ± 0.21 0.33 ± 0.05 3.03 ± 0.15 15.53 ± 0.31 

TUW-16 11.09 ± 0.22 0.37 ± 0.06 2.66 ± 0.13 14.34 ± 0.29 

Moraine boulders at Austurdalur (E Tungnahryggsjökull foreland) 

TUE-9 10.88 ± 0.22 0.16 ± 0.02 3.21 ± 0.16 15.81 ± 0.32 

TUE-10 10.75 ± 0.21 0.21 ± 0.03 3.07 ± 0.15 14.60 ± 0.29 

TUE-11 11.53 ± 0.23 0.16 ± 0.02 2.82 ± 0.14 14.03 ± 0.28 

TUE-12 11.90 ± 0.24 0.21 ± 0.03 2.70 ± 0.13 12.77 ± 0.26 

Glacially polished ridge separating Viðinesdalur and Kolbeinsdalur     

ELLID-1 7.20 ± 0.14 0.14 ± 0.02 6.96 ± 0.35 24.28 ± 0.49 

ELLID-2 7.01 ± 0.14 0.12 ± 0.02 7.18 ± 0.36 25.12 ± 0.50 

 1597 



ST8. 36Cl CRE dating results according to different 36Cl production rates from Ca spallation. 36Cl/35Cl and 35Cl/37Cl ratios were inferred from 1598 
measurements at the ASTER AMS facility. The numbers in italics correspond to the internal (analytical) uncertainty at one standard level. 1599 

Sample 

name 

Sample 

weight (g) 

mass of Cl in 

spike (mg) 

35Cl/37Cl 36Cl/35Cl (10-14) [Cl] in sample 

(ppm) 

[36Cl] (104 

atoms g-1) 

Age (yr) 

Licciardi et al. 

(2008) Ca 

spallation prod. 

rate 

Age (yr) 

Stone et al. (1996) Ca 

spallation prod. rate 

Age (yr) 

Schimmelpfennig 

et al. (2011) Ca 

spallation prod. 

rate 

Moraine boulders at Vesturdalur (W Tungnahryggsjökull foreland) 

TUW-9 82.98 1.823 27.225 ± 0.729 5.700 ± 0.357 3.5 0.46 ± 0.09 429 ± 99 (94) 453 ± 102 (99) 504 ± 118 (110) 

TUW-10 86.12 1.823 5.851 ± 0.144 0.878 ± 0.123 32.1 0.54 ± 0.10 384 ± 92 (88) 399 ± 94 (91) 437 ± 106 (100) 

TUW-11 88.96 1.826 5.198 ± 0.110 0.987 ± 0.136 41.0 0.73 ± 0.13 536 ± 119 (112) 554 ± 121 (116) 602 ± 135 (126) 

TUW-12 89.17 1.808 6.854 ± 0.091 1.324 ± 0.152 22.4 0.71 ± 0.10 567 ± 104 (95) 594 ± 105 (100) 654 ± 122 (110) 

TUW-13 87.33 1.846 5.850 ± 0.147 1.014 ± 0.152 32.0 0.65 ± 0.13 545 ± 129 (123) 565 ± 132 (128) 616 ± 148 (140) 

TUW-14 88.26 1.829 5.797 ± 0.192 2.044 ± 0.202 32.0 1.44 ± 0.17 1346 ± 211 (190) 1396 ± 213 (196) 1521 ± 245 (214) 

TUW-15 88.61 1.828 8.869 ± 0.150 1.317 ± 0.155 14.7 0.58 ± 0.09 560 ± 105 (98) 588 ± 107 (102) 649 ± 124 (113) 

TUW-16 88.50 1.825 4.402 ± 0.071 1.092 ± 0.164 67.0 1.19 ± 0.21 802 ± 187 (177) 823 ± 190 (181) 889 ± 210 (195) 

Moraine boulders at Austurdalur (E Tungnahryggsjökull foreland) 

TUE-9 93.53 1.814 13.846 ± 0.153 3.402 ± 0.272 7.2 1.33 ± 0.12 1283 ± 170 (141) 1357 ± 169 (149) 1509 ± 211 (165) 

TUE-10 85.51 1.831 12.715 ± 0.276 1.532 ± 0.172 9.0 0.60 ± 0.09 561 ± 102 (93) 592 ± 104 (98) 657 ± 122 (108) 

TUE-11 85.80 1.835 8.621 ± 0.110 3.620 ± 0.243 15.9 1.94 ± 0.15 1640 ± 201 (161) 1726 ± 197 (169) 1914 ± 248 (187) 

TUE-12 87.64 1.828 6.206 ± 0.088 3.172 ± 0.196 27.9 2.15 ± 0.16 1624 ± 197 (156) 1696 ± 193 (163) 1867 ± 238 (178) 

Glacially polished ridge separating Viðinesdalur and Kolbeinsdalur       

ELLID-1 28.88 1.864 68.854 ± 0.615 11.051 ± 0.438 3.5 12.438 ± 0.516 16256 ± 1685 (1217) 16921 ± 1618 (1262) 
18563 ± 2060 

(1380) 

ELLID-2 27.69 1.928 86.476 ± 0.876 10.014 ± 0.437 2.6 11.977 ± 0.547 16178 ± 1713 (1253) 16832 ± 1653 (1299) 
18460 ± 2090 

(1421) 

Blanks         Total atoms Cl Total atoms 
36Cl 

    

          (1017) (104)     

BK-1   1.800 297.029 ± 11.372 0.545 ± 0.097 2.941 ± 0.220 16.981 ± 3.034    

BK-2   1.818 279.307 ± 3.352 0.392 ± 0.079 3.251 ± 0.170 12.347 ± 2.484    

BK-3  1.888 328.059 ± 2.892 0.359 ± 0.069 2.650 ± 0.136 11.725 ± 2.249    

 1600 

 1601 



SUPPLEMENTARY FIGURES 1602 



SF1. Examples of Rhizocarpon geographicum group and Porpidia soredizodes thalli 1603 
measured over scaled field photos from the lichenometric station TUW-3. The lines refer to 1604 
the minimum circle (white) bounding the thalli outlines (yellow) and the diameter of the 1605 
circle (black). Note the contrasted size of the largest thalli in the different species. 1606 

 1607 



SF2. Location of the lichenometry stations in relation to the glacier snout position shown in 1608 
the different aerial photographs and the satellite image. The lines refer to the glacier margin 1609 
outlined over the aerial photo at a specific date (red) and at the previous date with aerial 1610 
photo available (green dashed line). The points display the location of the lichenometry 1611 
stations. The aerial photos provide the dating control for the validation of the lichenometric 1612 
dates. 1613 

 1614 

 1615 



SF3. Glacier surface reconstruction for the CRE- and lichenometry-dated glacial stages. 1616 

 1617 

 1618 



SF4. Examples of moraine boulders sampled for 36Cl CRE dating in Vesturdalur (“TUW” 1619 
samples) and Austurdalur (“TUE” samples) and their dates derived from the Licciardi et 1620 
al. (2008) 36Cl production rate from Ca spallation. 1621 

 1622 

 1623 



SF5. Location and BCE dates of the 36Cl samples from the Elliði crest. 1624 

 1625