Glacial geomorphology of Trygghamna, western Svalbard - Integrating terrestrial and submarine archives for a better understanding of past glacial dynamics

Glacial geomorphology of Trygghamna, western Svalbard - Integrating terrestrial and submarine archives for a better understanding of past glacial dynamics

Geomorphology 344 (2019) 75–89 Contents lists available at ScienceDirect Geomorphology journal homepage: Glacial g...

9MB Sizes 0 Downloads 3 Views

Geomorphology 344 (2019) 75–89

Contents lists available at ScienceDirect

Geomorphology journal homepage:

Glacial geomorphology of Trygghamna, western Svalbard - Integrating terrestrial and submarine archives for a better understanding of past glacial dynamics Nína Aradóttir a,⁎, Ólafur Ingólfsson a, Riko Noormets b, Ívar Örn Benediktsson a, Daniel Ben-Yehoshua c, Lena Håkansson b, Anders Schomacker d a

Institute of Earth Science, University of Iceland, Askja, Sturlugata 7, IS-101 Reykjavík, Iceland Department of Arctic Geology, University Center in Svalbard (UNIS), P.O. Box 156, N-9171 Longyearbyen, Norway c Svarmi ehf, Árleyni 22, IS-112 Reykjavík, Iceland d Department of Geosciences, UiT The Arctic University of Norway, N-9037 Tromsø, Norway b

a r t i c l e

i n f o

Article history: Received 6 May 2019 Received in revised form 11 July 2019 Accepted 12 July 2019 Available online 13 July 2019 Keywords: Surge-type glacier Geomorphological mapping Svalbard Glacier thermal regime

a b s t r a c t Detailed geomorphological mapping was carried out in the terrestrial and submarine forefields of Protektor-, Harriet- and Kjerulfbreen in Trygghamna, western Svalbard, based on high-resolution aerial images and bathymetric data. The mapping reveals that crevasse-squeeze ridges (CSRs) are only observed on land in the forefield of the surge-type Harriet- and Kjerulfbreen, and recessional moraines are only formed at the sea floor in relation to the retreat of these two glaciers. Different factors affect the preservation potential and formation of landforms between the two environments that could explain the absence of CSRs in the submarine environment and the terrestrial forefield of Protektorbreen. The landform assemblage in Trygghamna does not comply well with existing surge-type glacier landsystem models. We present a conceptual landsystem model for surge-type glaciers with combined terrestrial and marine margins, based on the geomorphological archive from Trygghamna. The contrast between the landform assemblages demonstrates how differences in the thermal regime result in different glacier behavior between the warm-based submarine margins and the inactive cold-based terrestrial margin. This study emphasizes the importance of integrating data from both archives to reconstruct past glacier behavior and understand the effect of different glacial dynamics and environments on the preservation potential of sediments and landforms. © 2019 Elsevier B.V. All rights reserved.

1. Introduction The Svalbard archipelago (60,667 km2; Fig. 1A), situated between 74° and 81°N, lies in the main transport path for North Atlantic air masses and ocean currents into the Arctic basin, which explains the relatively mild climate and sensitivity to climate change (Hanssen-Bauer et al., 1990; Dickson et al., 2000). Glaciers cover c. 57% of the land area (Nuth et al., 2013) and Svalbard is a hotspot for surging glaciers (e.g., Liestøl, 1969; Lefauconnier and Hagen, 1991; Hagen et al., 1993; Sevestre and Benn, 2015). The number of surge-type glaciers is still unknown with estimates ranging from 13% to 90% of all glaciers in the region (Lefauconnier and Hagen, 1991; Hagen et al., 1993; Jiskoot et al., 1998). Recent works assert that it is difficult to estimate the exact number of Svalbard surge-type glaciers as they are thought to have been more common during the Little Ice Age (LIA) than at present, based on their landform records (Liestøl, 1969, 1988; Hagen et al., 1993; ⁎ Corresponding author. E-mail address: [email protected] (N. Aradóttir). 0169-555X/© 2019 Elsevier B.V. All rights reserved.

Dowdeswell et al., 1995; Sevestre et al., 2015; Farnsworth et al., 2016; Lovell and Boston, 2017). After the termination of the last major glacier advance marking the peak of the LIA in Svalbard some 100 years ago, glaciers have undergone overall retreat and negative mass balance, exposing extensive areas of formerly glaciated landscape, in both terrestrial and submarine environments (Hagen et al., 1993, 2003; Nuth et al., 2013). The sedimentlandform assemblages at the recently deglaciated forefields in front of surge-type glaciers on Svalbard have been of interest for numerous studies of glacial dynamics and paleoclimate reconstructions (e.g., Boulton et al., 1996; Glasser et al., 1998; Evans, 2003; Ottesen and Dowdeswell, 2006; Ottesen et al., 2008; Flink et al., 2015, 2017; Streuff et al., 2018). However, only a few studies have integrated data from terrestrial and submarine forefields for a holistic view of the icemarginal environment and better understanding of glacial dynamics (Boulton, 1986; Kristensen et al., 2009; Farnsworth et al., 2017; Allaart et al., 2018; Lovell et al., 2018). When reconstructing glacial history and dynamics, surge-type glaciers usually pose a challenge because of their cyclic behavior,


N. Aradóttir et al. / Geomorphology 344 (2019) 75–89

Fig. 1. (A) Svalbard archipelago and the study area, Trygghamna. The box shows the location of Trygghamna, western Svalbard, on the archipelago. Place names referred to in the text are marked. Basemap: © Norwegian Polar Institute, 2017 ( (B) Aerial images from NPI (2009) and swath bathymetric dataset from the Norwegian Hydrographic Service (2000, 2007) were used to produce the geomorphological map. The frontal terminal moraine (dashed line) divides the submarine environment into the inner (shallower) and outer (deeper) part.

alternating between slow and rapid flow on a timescale of a few years to several decades (Meier and Post, 1969; Kamb et al., 1985; Raymond, 1987; Murray et al., 2003). Glacier surges are not considered a direct response to climate and are therefore less suitable for climate reconstructions, complicating the direct correlation between mass-balance driven glacier oscillations and changes in temperature and precipitation. This highlights the importance of identifying surge-type glaciers and understanding the reason behind individual glacier advances (Lefauconnier and Hagen, 1991; Yde and Paasche, 2010; Farnsworth et al., 2016). Landsystem models have been developed to help identify undocumented surge-type glaciers (Evans and Rea, 1999, 2003; Ottesen and Dowdeswell, 2006; Ottesen et al., 2008) and then modified to better depict the preservation potential and formation of sediments and landforms (Brynjólfsson et al., 2012, 2014; Schomacker et al., 2014; Brynjólfsson, 2015; Flink et al., 2015; Lønne, 2016). In the Trygghamna fjord, western Svalbard, the glaciers Protektor-, Harriet- and Kjerulfbreen previously drained into the fjord but now terminate almost exclusively on land (Fig. 1A). No glaciological studies have been conducted on these glaciers but a few have focused on the geomorphology. The geomorphology of the submarine forefield in Trygghamna indicates that the terminal moraine and associated debris lobes represent the maximum Holocene extent (Forwick, 2005; Forwick and Vorren, 2010; Streuff et al., 2018). The dynamic behavior of the glaciers is however debated. Recent studies suggest that at least some of the glaciers exhibited surge behavior based on terrestrial crevasse-squeeze ridges (CSRs) (Farnsworth et al., 2016; Wallin, 2016;

Ben-Yehoshua, 2017). However, Streuff et al. (2018) argued that they were not of surge-type based on the absence of overridden moraines and lineations in the submarine archive. The aim of this paper is to present a high-resolution geomorphological map of both the terrestrial and submarine forefields in Trygghamna and describe the landforms that occur in the forefields. This is in order to test if the glaciers have experienced surge behavior and to create a conceptual landsystem model for surge-type glaciers with integrated terrestrial and submarine margins. Geomorphological and historical data are further combined to reconstruct ice-marginal changes through time and for highlighting past glacial dynamics and thermal regimes. The geomorphological differences and similarities between the terrestrial and submarine forefields are emphasized, and controls on ice dynamics, processes, products and preservation potential in the different proglacial and ice-marginal environments are discussed. 2. Setting Trygghamna (78°14.5′N–13°51.0′E) is a 6 km long and 2 km wide fjord located at the outer part of northern Isfjorden. Five glaciers are situated in the fjord/valley today; Protektorbreen, Harrietbreen, Kjerulfbreen, Kiærbreen and Alkhornbreen (Fig. 1A; Table 1). Mt. Vӓrmlandryggen and Mt. Geologryggen, situated on the eastern side and head of the fjord, range from 310 to 574 m a.s.l., respectively, while mountain peaks from 330 to 849 m a.s.l. occur on the western side (Norwegian Polar Institute, 2003) (Fig. 1A). The western side of

N. Aradóttir et al. / Geomorphology 344 (2019) 75–89


Table 1 Description of the three larger glaciers located in the inner part of Trygghamna based on numbers from 2007 and 2017 (Koenig et al., 2010; Norwegian Polar Institute, 2003, 2017). Glacier


Protektorbreen Harrietbreen Kjerulfbreen

78°10.2′N 18°06.5′E 78°16.0′N 13°37.0′E 78°16.5′N 13°40.0′E

Area (km2)

Length (km)

~6.79 ~9.75 ~8

b3 ~3 b5

Trygghamna consists of low-grade metamorphic phyllite and the eastern side of deformed Mesozoic and Paleozoic rocks, sandstone and carbonates. The inner part of the fjord is rather shallow but with irregular sea-floor topography with depths down to 50 m b.s.l. Five islets, consisting of bedrock covered by debris, and 1–3 m deep shallows occur in this part of the fjord, usually in association with submarine transverse ridges. The seafloor in the outer part of the fjord is smoother at depths between ~100–200 m b.s.l. (Fig. 1B). Earliest efforts to map Trygghamna's bedrock geology and document ice-marginal positions were made in the late 19th and early 20th centuries. At that time, the large glaciers at the head of the fjord coalesced and reached significantly further into the fjord than at present, and the combined glacier was named Glacier Kjerulf (Nordenskiöld, 1892; Hamberg, 1905; Holtedahl, 1913; Isachsen, 1915; Liljequist, 1993) (Fig. 2A–B). Later observations revealed that the glacier had begun to retreat after the termination of the LIA (Fig. 2C), following the general trend in Svalbard (Hagen et al., 1993, 2003; Nuth et al., 2013). Ice-marginal reconstruction and historical data reveal that the terrestrial glacier margins were located adjacent to their terminal moraines between 1890 and 1909/10 and the tidewater margin was 0.5–0.6 km inside the submarine


Thermal regime

Large cirque Large cirque (partly grounded calving front) Outlet

Mostly cold-based Polythermal Polythermal

Flow direction NE E to ESE S to SE

terminal moraine (Fig. 3). The glaciers are thought to have been in overall retreat following the 1909/10 position. However, only six icemarginal positions were documented due to the resolution and timing of the satellite images and historical data. It is therefore acknowledged that minor readvances or longer halts in recession could have occurred between the documented ice-marginal positions. The overall retreat along the proposed major flow line of the individual glaciers from the terminal moraine up to 2016 is: Protektorbreen – 1.1 km; Harrietbreen – 2.4 km; Kjerulfbreen – 2.8 km (Fig. 3). Kjerulf- and Protektorbreen are completely terrestrial today but Harrietbreen has a partly grounded calving front. Protektor- and Harrietbreen originate from their own cirques and connect east of Knuvlen. Harriet- and Kjerulfbreen are attached at their confluence and form a single glacier front. Kjerulfbreen drains from the ice field on Geologpasset, which lies north of Trygghamna (Fig. 1A). 3. Methods Field campaigns were carried out in Trygghamna in the summers of 2015 and 2016. Landforms and sediments were measured, described,

Fig. 2. Historical data. (A) A panorama of Trygghamna where the glaciers are significantly larger than today, with a high ice front taken by Oscar Halldin in 1908 on a an excursion led by de Geer (1910). © Centrum för vetenskapshistoria, Kungl. vetenskapsakademien. (B) Topographical map, including ice-front positions in Trygghamna in 1909/10. The glaciers at the head of the fjord coalesced and were termed Glacier Kjerulf (Isachsen, 1915). (C) 1936 oblique aerial image of Trygghamna. The glacier margin is close to Protektorbreen's terminal moraine and the present coastline but Harriet- and Kjerulfbreen have an active, crevassed tidewater front. Note Knuvlen nunatak sticking out from the glacier surface (© Norwegian Polar Institute).


N. Aradóttir et al. / Geomorphology 344 (2019) 75–89

Fig. 3. Ice-marginal reconstructions for Trygghamna from 1909/1910 to 2016. The glacier has been in overall retreat since 1909/1910.

interpreted and classified according to their genetic origin. Descriptions of the sedimentological composition of landforms are according to Krüger and Kjær's (1999) data chart for glacial sediments. A handheld GPS device (Garmin GPSMAP 64s) was used to obtain the positions of landforms in the field. The geomorphology of the terrestrial forefield was mapped using orthorectified aerial images with pixel resolution between 0.4 and 0.5 m taken in 2009, acquired from the Norwegian Polar Institute (© Norwegian Polar Institute) (Fig. 1B). A digital elevation model (DEM) from 2009 with 5 m resolution, provided by NPI, was utilized to make contour lines and hillshades of the area. The Norwegian Hydrographic Service collected swath bathymetric data of the seafloor in Trygghamna in 2000 and 2007, using a Kongsberg EM-3000 multibeam echosounder. The grid size is 5 m in the outer most part of the fjord, 3 m at the distal side of the terminal moraine and 1 m in the inner part or proximal to the terminal moraine (Fig. 1B). No bathymetry data are available from the areas shallower than 1–2 m. The ice-marginal reconstructions (Fig. 3) are based on historical data (Nordenskiöld, 1892; Hamberg, 1905; Halldin, 1908; Holtedahl, 1913; Isachsen, 1915; Ohta et al., 1992; Liljequist, 1993), oblique aerial images from 1936 (NPI) (Fig. 2A–C), satellite images (LANDSAT) and GLIMS Glacier Database (Koenig et al., 2010). Analysis of the data and mapping was conducted using ESRI ArcGIS 10.4 and 10.3 and georeferenced when needed. In addition,

for the terrestrial geomorphological mapping, the data were viewed in stereoscopic view using the ERDAS IMAGINE 2015 with the SAFA extension for ArcMap and the bathymetric data were viewed in the free software iView4D in 3D. The spatial reference system used for all data handling is WGS84/UTM 33N. The zoom level while mapping was set to 1:1500, for both the terrestrial and submarine environments. The terrestrial mapping is done by combination of field work and remote sensing, as suggested by Chandler et al. (2018). However, the submarine mapping was solely based on the bathymetric data as no acoustic profiles or sediment cores exist from the inner part of Trygghamna (Dowdeswell et al., 2016; Batchelor et al., 2017). As the intended final product was a composite map of the terrestrial and submarine forefields, the same mapping style was applied in both settings. However, slightly different terminology and symbols were used between the two different environments. The final design and production of the map was carried out in Canvas X. 4. Mapped landforms A composite geomorphological map (Fig. 4), including the terrestrial and submarine forefields of Kjerulf-, Harriet-, and Protektorbreen, has been produced for the inner part of Trygghamna. As the focus is on the recently deglaciated forefields, the area beyond

N. Aradóttir et al. / Geomorphology 344 (2019) 75–89


Fig. 4. Composite terrestrial and submarine geomorphological map of the forefields of Protektor-, Harriet- and Kjerulfbreen in the inner part of Trygghamna. Rose diagrams (adapted from Ben-Yehoshua, 2017) show the orientation for CSRs in the forefield of Harriet- and Kjerulfbreen. Terrestrial contour lines are 20 m and bathymetric 10 m).

the three glacier forefields is mapped as an extra-marginal surface in the terrestrial environment and distal glaciomarine sediments in the submarine environment. The geomorphology is described on the basis of the classification of landforms into; subglacial, supraglacial, glaciofluvial and ice-marginal. In addition, the sedimentary composition of key terrestrial landforms is described.

4.1. Subglacial landforms 4.1.1. Fluted till plain and ice-moulded bedrock In this study, the terrestrial subglacial surface includes fluted till plain and debris covered and exposed bedrock, glacially or fluvially moulded. The surface cover in the glacier forefields is composed of


N. Aradóttir et al. / Geomorphology 344 (2019) 75–89

lithologies. Angular to sub-angular phyllites appear to be more common in the forefield of Protektorbreen (Fig. 5A–C) than in the forefields of Harriet- and Kjerulfbreen, where rounded sandstone clasts appear more common (Fig. 6A–D). Bedrock exposures are numerous (Figs. 5A–B and 6E) and the sediment cover is the most discontinuous in the forefield of Protektorbreen (Fig. 4). Polished surface and striations occur on a few bedrock outcrops in the forefields of Protektor- and Kjerulfbreen. In the forefield of Protektorbreen, the average direction of striae is WSW-ENE (250°– 70°) but two sets of directions were observed in the forefield of Kjerulfbreen, ~NW-SE (300°–120°) and N-S (8°–188°). The orientation of the striae correlates well to the present day glacier flow for Protektorbreen (SW-NE) and one of the sets from Kjerulfbreen (NW-SE). The N-S set of striae in the forefield of Kjerulfbreen is considered to pre-date the LIA, from a time when the whole area was covered by a glacier (Salvigsen et al., 1990; Forwick and Vorren, 2010), flowing down through the valley occupied by Kiærbreen and Lovénvatnet today. Flutes are common on the till plain in the terrestrial forefields of the three glaciers. Most have vague appearance in the field and can primarily be observed on the aerial images. Around 60 prominent flutes have been marked on the map (Fig. 4). The flutes range from 1 to 97 m in length, are up to 1 m wide and with a relief b0.3 m, based both on field evidence and remote sensing. They consist of till, similar to the surrounding till plain, except for seemingly smaller clast size. Some of the flutes have a boulder at their head (Figs. 5C and 6A). The orientation and abundance of the flutes varies between the forefields, reflecting the former direction of ice-flow. In the forefield of Protektorbreen, the flutes are generally subtle and with average orientation of WSW-ENE (250°–70°) (Fig. 5C). The flutes are, however, much more prominent in the forefield of Harrietbreen where around 40 were mapped. Their main orientation is W-E (270°–90°) but a few of them curve slightly northwards, following the local topography. Some cross-cut debris bands or are draped by CSRs (Fig. 6A). In the forefield of Kjerulfbreen, only one flute was observed in the field despite the aerial images clearly showing a streamlined surface with sub-parallel lineations. Their main orientation is roughly NW-SE (310°–130°). Flutes are thought to form during ice-bed coupling and subglacial deformation of till (Boulton, 1976; Kjær et al., 2006; Ives and Iverson, 2019). The flutes are easily degraded by water and wind and therefore have low preservation potential, explaining their often obscure appearance in the field. They can form in front of both surging and non-surging glaciers and thus cannot be used independently as a diagnostic landform for surge-type glaciers (Evans and Rea, 1999, 2003).

Fig. 5. The forefield of Protektorbreen. The large arrows indicate the glacier flow direction. (A) Seen from the lateral terminal moraine, debris covered bedrock hill, dissected by inactive fluvial channels in the middle of the figure with kames and kettles below it. The trimline on Knuvlen in the background. (B) Aerial image (NPI, 2009) showing the medial moraine (white dashed line) that marks the former intersection of Protektorand Harrietbreen. Few CSRs are located on Protektorbreen's side. Supraglacial debris bands (black dashed lines), frequently exposed bedrock and high amount of fluvial activity are characteristic for Protektorbreen's forefield. (C) A vague flute in the forefield of Protektorbreen. Note high frequency of sub-angular phyllite.

clast-rich till. Generally, the till is coarse-grained with sub-angular to sub-rounded gravel- to boulder-sized clasts. Light and brownish sandstones, carboniferous shales and phyllites are the most frequent

4.1.2. Crevasse-squeeze ridges (CSRs) A dense network of CSRs covers most of the fluted till plain in the terrestrial forefields of Harriet- and Kjerulfbreen. Around 1400 CSRs were mapped in total (Fig. 4). The CSRs are usually orientated transverse or slightly oblique to the former ice-flow direction of the individual glaciers. In the field, they range from 0.5 to 100 m in length, with an average around 15 m, and are usually about 1 m wide and 0.5 m high, but examples of 7 m wide and 2 m high ridges do occur. The CSRs consist of a compact, clast-rich, matrix-supported diamict (Ben-Yehoshua, 2017) (Fig. 6D), and they rest directly on either the till plain, beach gravel, flutes, or bedrock (Fig. 6A–E). In the forefield of Harrietbreen b400 CSRs were mapped, most of them aligned N-S (Fig. 4). Usually they occur as sole features (Fig. 6B, C, E) but CSRs intersecting each other and flutes were also observed (Fig. 6A). The network of CSRs is denser in the forefield of Kjerulfbreen where over 1000 ridges were mapped. Here, the CSRs often seem to intersect, forming complex patterns with variable orientation. The dominating orientation is, however, NNESSW (Fig. 4). They occur on the till plain close to the coast and mid-way to the glacier, where they become less frequent and are

N. Aradóttir et al. / Geomorphology 344 (2019) 75–89


Fig. 6. The forefield of Harriet- and Kjerulfbreen. The large arrows indicate direction of glacier flow. (A) Flute and CSR cross-cutting in the forefield of Harrietbreen. No obvious difference could be detected between the two landforms, except for their orientation. (B) CSRs orientated transverse to the glacier front in the forefield of Harrietbreen. (C) CSRs by the bay in the forefield of Kjerulfbreen. (D) A cross-section of a CSR in the forefield of Kjerulfbreen. The internal composition consisted of compacted, clast-rich, matrix supported diamict. (E) The forefield of Kjerulfbreen seen from the fjord. Fluted till plain with CSRs located by the fjord, dissected by inactive fluvial channels exposing bedrock to the left. Hummocky moraine is prominent in the background with solifluction lobes and higher amount of vegetation but no flutes present.

then absent closest to the glacier front (Figs. 4 and 6E). The CSRs are almost absent in the forefield of Protektorbreen. However, twelve CSRs with N-S orientation have been mapped on Protektorbreen's side close to the medial moraine that separates the two forefields (Fig. 5B). CSRs have been used as one of the main diagnostic criteria for the identification of surge-type glaciers (Sharp, 1985; Evans and Rea, 1999, 2003; Schomacker et al., 2014; Farnsworth et al., 2016). Their formation is generally thought to be by upward infilling of

saturated sediments into basal crevasses that form in association with the ongitudinal and extensional stress during a surge. The process happens towards the termination of the surge and subsequent meltout (Rea and Evans, 2011; Lovell et al., 2015). The complex network of CSRs in the forefield of Kjerulfbreen indicates that the base of the glacier was heavily crevassed in multiple directions. Their preservation potential is usually high in front of surge-type glaciers because of the stagnant glacier front that down wastes in situ after a surge (Evans and Rea, 1999, 2003).


N. Aradóttir et al. / Geomorphology 344 (2019) 75–89

4.2. Supraglacial landforms 4.2.1. Hummocky moraine Hummocky moraine is located on the proximal side of the lateral terminal moraine in the terrestrial forefield of Kjerulfbreen, dominated by small, irregular hummocks (1–5 m in height), small streams and lakes and more abundant vegetation than on the fluted till plain (Fig. 4). The surface cover is generally a clast-rich diamict with coarse, sub-angular to sub-rounded clasts. Zones with sorted gravel to sand and clay occur sporadically. Striations were observed on some of the clasts, but no flutes or CSRs are present (Fig. 6E). Marine shell fragments and occasional whole shells frequently occur on the surface. Boulders are located at the crest of hills or in depressions due to fall sorting and backslumping are associated with the hummocks and sediment lobes on slopes. Flatter areas with waterlain sediments and denser vegetation are located between the hummocks (Fig. 7A). The morphology of the hummocky moraine is the result of a differential dead-ice melting below a variably thick debris cover (Boulton, 1968; Hambrey et al., 1997; Kjær and Krüger, 2001; Schomacker and Kjær, 2008) and can be modified substantially over time due to degradation (Schomacker and Kjær, 2007, 2008). The wide range of material encountered in the hummocky terrain is interpreted to be reworked, and the hummock formation has been associated with thrusting, where material has been brought into en- and supraglacial positions (Sharp, 1985, 1988; Hambrey et al., 1997; Benediktsson et al., 2010). Stagnation of the glacier snout, following a surge-event, commonly leads to the formation of hummocky moraine on the inside of a glaciotectonic terminal moraine (Evans and Rea, 2003; Benediktsson et al., 2008; Schomacker et al., 2014). Hummocky terrain does however appear in front of both surging and non-surging glaciers and can thus

not be used independently as diagnostic for surge-type glaciers (Evans and Rea, 1999, 2003). 4.2.2. Kames and kettles Kames and kettles characterize the area north of the lateral moraine in the terrestrial forefield of Protektorbreen and south of Lovénvatnet in the terrestrial forefield of Kjerulfbreen (Fig. 4) with small, gentle depressions and hummocks consisting of a mix of sand and rounded to sub-angular boulders. Some of the depressions are filled with water and inactive fluvial channels are common (Fig. 5A). Kame terraces mark the western boundary of the area in the forefield of Kjerulfbreen. Examination of the aerial images from 1936 reveals an ice-marginal lake in front of Kjerulfbreen, agree with the location of the kame terraces. Kettle holes, with an almost rectangular shape, are present at the fluted subglacial till plain on the peninsula in front of Kjerulfbreen (Fig. 4). Their sides are covered with coarser material than the surrounding till plain and have no surface cracks. 4.2.3. Supraglacial debris bands and medial moraine Supraglacial debris bands consisting of angular to sub-angular cobbles of phyllite, orientated approximately in the ice-flow direction, are present in all three terrestrial forefields (Fig. 4). The bands have the highest frequency in the forefield of Protektorbreen and close to the margins of Harriet- and Kjerulfbreen. The bands are around 0.3–1 m wide and can be up to 500 m long (Figs. 5B and 7B). They can often be traced to supraglacial debris cover on the glacier, which was observed to have the same composition. Interaction medial moraines are present at the intersection between Harriet- and Kjerulfbreen and at the eastern margin of Kjerulfbreen.

Fig. 7. Examples of terrestrial landforms. (A) Backslumping, fall sorting and waterlain sediments are characteristics for the hummocky moraine in the forefield of Kjerulfbreen. (B) Supraglacial debris bands and eskers in the forefield of Protektorbreen. The lateral terminal moraine is to the right. (C) Esker with a sinuous shape, eroded by fluvial activity at the southern margin of Protektorbreen. (D) A cross section of an esker in the forefield of Protektorbreen. It consists of horizontal layers of sorted sand to gravel, mostly sub-rounded to rounded.

N. Aradóttir et al. / Geomorphology 344 (2019) 75–89

There are three parallel supraglacial debris bands at the joint of the forefields of Protektor- and Harrietbreen, right below Knuvlen (Fig. 4). The debris bands consist of angular to sub-rounded sandstone and phyllite and have a W-E orientation (Fig. 5B). The debris bands between the forefields of Harriet- and Protektorbreen are the remnant of a melted out interaction medial moraine formed by the lateral compression caused by the confluence of the two glaciers. The shape of the debris bands is dependent on the relative strength of the flow units (Eyles and Rogerson, 1978; Hambrey et al., 1999), indicating that flow from both glaciers affected that area to some extent. However, since the orientation of the debris bands correlates well with the ice-flow direction of Harrietbreen, it suggests that Harrietbreen was more active than Protektorbreen during their last advance. The preservation potential for medial moraines and supraglacial debris is low and therefore not always easy to distinguish in the field after the glaciers have retreated (Glasser and Hambrey, 2003). 4.3. Glaciofluvial landforms 4.3.1. Channels Numerous channels are observed, both active and inactive, in all three forefields (Fig. 4). Individual inactive channels are usually orientated transverse to the glacier front or follow the topography (Fig. 5B). Bedrock is frequently exposed in association with the channels (Figs. 5A–B and 6E). The seasonal or inactive channels are interpreted to be lateral meltwater channels. Lateral meltwater channels often form at the margins of polythermal or cold based glaciers, when water cannot penetrate through the frozen bed. Their morphology can further be used to trace the retreat of the glacier margin (Dyke, 1993; Ó Cofaigh et al., 1999; Greenwood et al., 2007). 4.3.2. Outwash plain The surface material of the outwash plains consists of sub-rounded to rounded, sorted sand and gravel and has gently down-glacier sloping surfaces. The outwash plain in the forefield of Protektorbreen is the most extensive and covers about 50% of the forefield (Figs. 4 and 5B). In the forefield of Harriet- and Kjerulfbreen, outwash plains are less dominant and limited to restricted areas in front of the glacier margin or in association with channels (Figs. 4 and 6E). As a result of the difference in the extent of outwash plains between the forefields, fluvial erosion is more dominant in the forefield of Protektorbreen. The preservation potential of landforms would therefore be expected to be lower in the forefield of Protektorbreen due to this difference (Brynjólfsson et al., 2014; Brynjólfsson, 2015). 4.3.3. Eskers Several sinuous eskers appear in the forefields of Protektor- and Harrietbreen (Fig. 4). They vary between 30 and 60 m long, 1–2 m wide and 0.5–2 m high. The surface cover appears to have a finer matrix than the surrounding diamict and cracks are occasionally visible on the surface, indicating dead-ice melting below the deposits, which suggests either en- or supraglacial origin (Evans and Twigg, 2002). Small sections in the ridges reveled that they are composed of horizontally layered silt to medium sand and clast-supported, fine gravel to 30 cm boulders, with sub-angular to rounded clasts (Fig. 7C–D). The ones in the forefield of Protektorbreen are orientated roughly SW-NE and W-E, depending on their location but in the forefield of Harrietbreen they are all situated on a gentle slope with an orientation of roughly N-S (Fig. 4). 4.4. Ice-marginal landforms 4.4.1. Lateral and frontal terminal moraines Lateral terminal moraines are located adjacent to the mountain sides in the terrestrial part of the forefields of Protektor- and Kjerulfbreen (Fig. 4), considered to represent the LIA maximum extent of the glaciers, based on the general record from the surrounding fjords (Werner, 1993;


Svendsen and Mangerud, 1997). The proximity of the ~1900 glacier margin to the terminal moraines further supports this interpretation. The lateral terminal moraine, with roughly W-E orientation, delimits the southern margin of Protektorbreen's forefield. The moraine ridge is up to 60 m high, over 1 km long and up to almost 300 m wide (Fig. 7B–C). The surface cover consists of a clast rich till with subangular to sub-rounded phyllite and sandstone clasts of cobble to boulder size. The matrix is a mixture of sand, silt and clay but most fines have been washed away from the surface. The northern and middle part of the moraine is hummocky, with cracks and sediment lobes present. In the terrestrial forefield of Kjerulfbreen, the lateral terminal moraine lies adjacent to the mountain side and the southern side of Lovénvatnet. It is orientated NW-SE in the proximal part but curves towards the E next to Lovénvatnet where it fades into a lower relief ridge. The moraine is over 2 km long, up to 70 m wide and 30 m high. The surface cover consists of a diamict, with smaller grain size and lower frequency of angular clasts than the moraine in the forefield of Protektorbreen and higher frequency of finer matrix is visible on its surface. Few cracks were observed along its side. The larger grain size and higher frequency of angular clasts on the Protektorbreen moraine indicate that it is to some extent formed by dumping and rockfall from the adjacent mountain sides (Benn and Evans, 2010). The hummocky surface, cracks, solifluction lobes and the large dimension all indicate that these ridges are partly ice-cored and that dead-ice melting has occurred (Schomacker and Kjær, 2008). This partly explains the large size of the moraine in front of Protektorbreen. This is similar to large thrust moraines with ice-cored proximal slopes that have been described in front of surge-type glaciers in Svalbard and Iceland (e.g., Boulton et al., 1999; Evans and Rea, 1999, 2003; Benediktsson et al., 2010, 2015). In the submarine forefield, a large NE-SW orientated frontal terminal moraine almost crosses the fjord transverse to its axis (Fig. 4). It is interpreted to be a frontal terminal moraine as a direct continuation of the terrestrial lateral terminal moraine onto the seafloor and also represents the LIA maximum extent of Harriet- and Kjerulfbreen (Forwick, 2005; Forwick and Vorren, 2010; Streuff et al., 2018). The ridge is located 2.4–2.8 km from the 2016 margins of Harriet- and Kjerulfbreen and 0.5–0.6 km from the 1909/10 terminal position (Fig. 3). The frontal terminal moraine is semi-continuous and can be divided into three main parts, with different morphologies. The NE part is ~350 m long and up to 150 m wide and has a slightly asymmetric profile, ~10 m high on the proximal side with a slope of 13–19° but ~25 m high on the distal side with a slope of 10–12°. The terminal moraine is mostly single crested and has a crescent shape. The NE-part of the ridge is connected to a topographic high that extends into a recessional moraine on the proximal side of the terminal moraine. The middle part is the longest, ~900 m, and up to 150 m at the widest part and consists of three lobes, two large-scale with a crescent shape and one smaller with straighter shape. The middle part varies from ~5–20 m high and has an asymmetric profile, with a steeper proximal slope (17–24°) and a shallower distal slope (10–13°). The SW part of the ridge is the smallest, ~500 m long, up to 40 m wide and 10 m high. This section has a symmetric cross-profile, 20–30° and is curvilinear in planform. The water depth is shallowest over the NE part, ~6–10 m, but down to ~30 m in the middle of the fjord (Figs. 8B and 9B). The formation of the frontal terminal moraine is connected to bulldozing of marine muds and subglacial sediments in front of an advancing glacier terminus and/or by sediment accumulation during periods of longer terminal still stands (Solheim, 1991; Boulton et al., 1996; Kristensen et al., 2009). Similar submarine terminal moraines have been described from both non-surging and surge-type glaciers in Svalbard (Plassen et al., 2004), where the latter are often multicrested (e.g., Solheim, 1991; Boulton et al., 1999; Ottesen and Dowdeswell, 2006; Ottesen et al., 2008; Flink et al., 2015, 2017). The orientation of the ridge, difference in dimension and non-continuous SW part is probably because Harriet- and Kjerulfbreen were more active


N. Aradóttir et al. / Geomorphology 344 (2019) 75–89

N. Aradóttir et al. / Geomorphology 344 (2019) 75–89

during the time of its formation, influencing the ridge morphology. In addition, based on the low frequency of sediments in the terrestrial forefield of Protektorbreen, the glacier had limited access to sediments compared to the other glaciers.

4.4.2. Recessional moraines Sub-parallel transverse ridges, interpreted as recessional moraines, are prominent in the submarine forefield. Their orientation is generally transverse to the fjord axis (NE-SW) but differs slightly around topographic highs and close to the shore (Fig. 4). Over 300 closely spaced, sub-parallel ridges have been mapped. They are 7–760 m long (average of 124 m), ~5–20 m wide and ~0.5–4 m high and are curvilinear in planform. Few of them appear to be linked together but no rhombohedral pattern is observed (Figs. 4 and 8A–B). Their cross-profile varies greatly, from symmetric to asymmetric with both steeper proximal and distal side. The slope angle is also highly variable (6–20°) (Fig. 9A–B). Several segments of moraines with larger dimension were mapped, ranging from 33 m to over 1000 m in length, ~20–50 m in width and ~4–12 m in height. They constitute almost a continuous line crossing the fjord in two places inside the terminal moraine (Figs. 8A and 9A). The larger moraines correlate well with the 1909/10 and 1936 ice marginal positions (Fig. 3). Two recessional moraines, 40 m and 90 m long and up 20 m wide occur close to the shore in the terrestrial forefield of Harrietbreen, transverse to the glacier margin (Fig. 4), indicating that the tidewater margin of Harrietbreen was depositing or pushing material. They could not be accessed in the field and are surrounded by fluvial outwash plain and partly eroded (Fig. 5A). Recessional moraines are thought to be formed by pushing during small winter readvances (e.g., Boulton, 1986; Solheim, 1991; Bennett, 2001; Ottesen and Dowdeswell, 2006; Ottesen et al., 2008; Flink et al., 2015) and have been described from other tidewater glaciers, both surging and non-surging, in Svalbard. Submarine CSRs and De Geer moraines have often similar morphology as recessional moraines and the problem in differencing between them has been discussed (Lindén and Möller, 2005; Ottesen et al., 2008; Bouvier et al., 2015; Burton et al., 2016). Thus, without any view of their internal architecture, we leave it open for debate whether the ridges could also be formed by the squeezing of soft subglacial sediments into basal crevasses (Streuff et al., 2018). Several factors could explain the size difference of these recessional moraine ridges. The larger ridges could have been formed during periods of longer still-stands with sustained sediment delivery and deformation around the same position. Small readvances may have contributed further by incorporating former ridges into the larger one. The correlation of the large ridges with shallows and islets suggests these could have acted as pinning points providing a temporary stability for the ice front during retreat (Maclachlan et al., 2010; Sund et al., 2011; Burton et al., 2016; Allaart et al., 2018). The regional geology and topography thus play an important role for the glacier retreat and the morphological development of their forefields. The pattern of retreat is further preserved in the ice-marginal and geomorphological record. Recessional moraines can sometimes be connected to annual winter positions of the glacier front and have therefore been termed annual moraines (Boulton, 1986; Evans and Twigg, 2002; Ottesen and Dowdeswell, 2006; Ottesen et al., 2008). It cannot always be proven that they formed annually as it is often difficult to connect them to former positions of the glacier front due to lack of historical data, low preservation potential and their incompleteness (Jónsson et al., 2014; Flink et al., 2015). The 1909/10 and 1936 margins coincide well with the large-scale retreat moraines (Fig. 3) but only


twelve to eighteen ridges are located between them. It can therefore not be established that they formed annually in this setting. 4.4.3. Trimline Trimlines can be observed on the SE and NE sides of Knuvlen at 140– 280 m a.s.l. and 60–100 m a.s.l., respectively and north of Kjerulfbreen at 180–280 m a.s.l (Fig. 4). The slope below the trimline has fresher appearance but is more weathered above it (Fig. 5A). They are considered to represent the LIA maximum extent of the glaciers and demonstrate the vertical loss of ice since then (Fig. 2A–C). 4.4.4. Debris flow apron A large, lobate accumulation of sediments is located at the distal side of the submarine frontal terminal moraine, interpreted to be a glacigenic debris-flow apron formed during the last maximum extent of the glaciers (Forwick, 2005; Streuff et al., 2018). The lobe initiates at ~30–40 m b.s.l. and reaches down to ~100 m b.s.l. with a maximum length of ~0.7 km and width of 1.6 km. The lobe has a gentle slope, which flattens out towards the bottom. The surface is irregular with blocks and mounds and several superimposed smaller lobes. The lobe gets less extensive towards the SW (Figs. 8B and 9B). Smaller and less distinct lobes are located at the distal side of couple of the larger ridges, up to 60 m long (Fig. 4). The formation of the apron is the result of a combination of ice push and quasi-continuous slope failure of glacigenic material brought subglacially by the glacier to the terminal moraine (Ottesen et al., 2008; Kristensen et al., 2009). Similar lobes have been described from other fjords in Svalbard in association with terminal moraines, at both surging and non-surging glaciers (Plassen et al., 2004; Ottesen and Dowdeswell, 2006; Ottesen et al., 2008; Flink et al., 2015, 2017). The superimposed lobes indicate high sediment input and could be related to separate surge-events (Flink et al., 2017). The smaller lobes at the large-scale recessional moraines are also interpreted to be debris-flow lobes, supporting that they formed during longer still-stands or small readvances of the glacier. 5. Discussion 5.1. Surge evidence in Trygghamna Detailed geomorphological mapping of the terrestrial and submarine forefields of Protektor-, Harriet- and Kjerulfbreen in Trygghamna, reveals clear contrasts between the terrestrial and submarine environments. None of the landform assemblages are strictly indicative of former surge behavior based on published landsystem models (Evans and Rea, 1999, 2003; Ottesen and Dowdeswell, 2006; Ottesen et al., 2008; Brynjólfsson et al., 2012, 2014; Schomacker et al., 2014; Brynjólfsson, 2015; Flink et al., 2015). However, the results from this study combined with recent studies in the area (Wallin, 2016; BenYehoshua, 2017), suggest that Harriet- and Kjerulfbreen exhibited surge behavior in the past, based particularly on the presence of terrestrial CSRs. This also agrees with a remote sensing survey, utilizing the presence of CSRs to identify undocumented surge-type glaciers (Farnsworth et al., 2016). Fluted till plain, CSRs, hummocky moraine, kame and kettle topography and en- and supraglacial debris bands all conform to the surge-type outlet or cirque glacier models (Evans and Rea, 1999, 2003; Brynjólfsson et al., 2012, 2014; Schomacker et al., 2014; Brynjólfsson, 2015). Despite the few CSRs that were observed on the northern fringe of the forefield of Protektorbreen, the overall landform assemblage is not conclusive of past surge behavior, although it can be argued that the glacier surged in the past based solely on the presence of CSRs (Farnsworth

Fig. 8. Examples of submarine landforms. The black arrows indicate the generalized direction of glacier flow and the black lines location of seafloor profiles (Fig. 9A–B). (A) Recessional moraines. Note that the moraines often correlate with the islets and shallows. (B) Large frontal terminal moraine (dashed line) with debris flow apron on its distal side and recessional moraines on the proximal side.


N. Aradóttir et al. / Geomorphology 344 (2019) 75–89

Fig. 9. Longitudinal profiles of submarine landforms. (A) Fig. 8A. (B) Fig. 8B. Arrows point at recessional moraines.

et al., 2016). Because of their location close to Harrietbreen's forefield, these CSRs could be the result of intense lateral shearing and crevassing related to differences in ice-flow velocities between the two glaciers during past surges of Harrietbreen. Other landsystem elements in the forefield of Protektorbreen, such as extensive outwash and fluvial erosion, abundance of en- and supraglacial debris, and kame and kettle topography could all conform to both the surge-type valley and cirque glacier landsystem models (Brynjólfsson et al., 2012, 2014; Brynjólfsson, 2015) as well as to the non-surging polythermal glacier landsystem model (Glasser and Hambrey, 2003). Therefore, based on this landform assemblage along with the lack of historical and glaciological evidence for surging behavior, it cannot be determined if Protektorbreen surged in the past. The absence of diagnostic surge-type landforms in the terrestrial archive could owe to generally coarse and thin cover of subglacial sediments (Lønne, 2016), intense fluvial erosion, high amount of supraand englacial debris (Brynjólfsson et al., 2012, 2014; Brynjólfsson, 2015) and degradation due to permafrost thawing (Schomacker and Kjær, 2008). These characteristics all correspond well with the landform assemblage in the forefield of Protektorbreen and we propose that they could be the reason for the absence of certain landforms, such as CSRs and concertina eskers. Streuff et al. (2018) suggested that the absence of glacial lineations and overridden moraines in the submarine geomorphology indicated relatively slow ice flow in Trygghamna during the latest advance. However, they did not include the terrestrial archive and thus suggested this based on the submarine archive alone. The lack of CSRs, streamlined lineations and eskers in the submarine landform assemblage shows little correspondence with published surge-type tidewater models and could thus be taken to indicate non-surging behavior (Ottesen and Dowdeswell, 2006; Ottesen et al., 2008; Flink et al., 2015; Streuff et al., 2018). As Harriet- and Kjerulfbreen are considered to have been more active than Protektorbreen during the latest advance, the geomorphology in the submarine environment is dominantly shaped by their activity. However, when combined with the terrestrial forefields, which strongly indicate surge behavior, we interpret the marine based part of the glacier must have undergone surge as well. In the submarine archive, CSRs can have relatively low relief (b1 m) and are therefore likely to be draped and eventually buried by glaciomarine sediments with time (Plassen et al., 2004; Flink et al., 2015, 2017; Fransner et al., 2017). High sedimentation rates have been estimated in outer Trygghamna due to comparatively high precipitation and low-grade metamorphic rock that is relatively easily eroded (Plassen et al., 2004; Forwick, 2005; Forwick and

Vorren, 2010), which supports the hypotheses that CSRs might have been formed but not preserved. CSRs are often preserved proximal to the terminal moraine, indicating ice stagnation for some time after a surge (Solheim, 1991; Ottesen and Dowdeswell, 2006; Ottesen et al., 2008; Flink et al., 2015). In Trygghamna, the recessional moraines right at the proximal side of the terminal moraine and lack of CSRs suggests that the marine-based part was never stagnant and the small winter readvances during the overall retreat could therefore have overprinted any potential CSRs formed at the end of a surge. This shows that caution must be exercised when interpreting the lack of landforms. 5.2. Landsystem model for surge-type glaciers with composite terrestrial and marine margins In contrast to the published landsystem models that are based on more simplified settings (e.g., Evans and Rea, 1999, 2003; Ottesen and Dowdeswell, 2006; Ottesen et al., 2008; Brynjólfsson et al., 2012, 2014; Schomacker et al., 2014; Brynjólfsson, 2015; Flink et al., 2015), a conceptual landsystem model from Trygghamna integrates the terrestrial and submarine geomorphological archives of surge-type glaciers (Fig. 10). The terrestrial environment consists of CSRs on a fluted till plain, sinous esker, supra- and englacial sediments that is dissected by an outwash plain. Lateral terminal moraine delimits the forefield, with hummocky moraine and kames and kettles on the ice proximal side. The submarine environments contain recessional moraines, often in relation to bedrock islets, on the ice proximal side of a frontal terminal moraine with a debrisflow apron on the ice distal side. The model reflects differences in geomorphology between the terrestrial and submarine environments of surge-type glacier forefields and demonstrates differences in glacial dynamics and thermal regimes within these two environments during a surge and subsequent retreat. This difference in the thermal regime during the retreat is vital when investigating geomorphological evidences for surge behavior as it has a great effect on the preservation potential of landforms as outlined above. Previous studies on surge-type glaciers have concluded that the terrestrial margins are inactive and cold-based during the retreat, and that they waste down in situ, resulting in high preservation of landforms (Evans and Rea, 1999, 2003). The lack of recessional moraines, high preservation of CSRs and lateral meltwater channels in the terrestrial forefields are all supportive of that. On the other hand, surge-type marine margins are considered to undergo small winter readvances during the overall retreat based on the occurrence of recessional moraines

N. Aradóttir et al. / Geomorphology 344 (2019) 75–89


Fig. 10. Conceptual landsystem model for surge-type glaciers with composite terrestrial and marine margin in Svalbard, based on observations from Trygghamna, Svalbard. Terrestrial environment: CSRs, flutes, supra- and englacial sediment and sinous esker characterize the subglacial till plain. Fluvial outwash plain, containing active and inactive channels, erodes parts of the till plain. Lateral terminal moraine delimits the forefield with hummocky moraine on the proximal side and kames and kettles. Submarine environment: Transverse recessional moraines, often in correlation with bedrock islets, are located proximal to the frontal terminal moraine and adjacent debris flow apron.

(Boulton, 1986; Solheim, 1991; Ottesen and Dowdeswell, 2006; Ottesen et al., 2008; Flink et al., 2015). Flink et al. (2015) explained the winter readvances by longitudinal extension at the terminus together with low calving rates due to suppression from sea ice. The recessional moraines in the submarine archive supports that the glacier margin was active during the retreat and furthermore could explain the absence of CSRs as previously stated. This contrast between the environments suggests fundamental differences in the glacial dynamics between the inactive cold-based terrestrial margin and the active retreat of the warmbased submarine margin. Our results agree with Allaart et al. (2018) interpretation of terrestrial and submarine landforms at the non-surging Nordenskiöldbreen in Svalbard. Our mapping of the terrestrial and submarine forefields emphasizes the importance of integrating both parts when possible in order to enable holistic assessments of glacial geomorphological archives and reconstructions of past glacier behavior. Tidewater glaciers drain twothird of the glaciated area in Svalbard and due to recent retreat (Nuth et al., 2013), combined terrestrial and submarine forefields are becoming more common. The reduction in annual sea-ice allows for more comprehensive collection of bathymetric data from recently deglaciated areas around the archipelago (Serreze et al., 2003; Ottesen et al., 2017), which increases the opportunities to combine terrestrial and submarine data for better understanding the dynamic evolution of Svalbard's glacial environments (Kristensen et al., 2009; Farnsworth et al., 2017; Allaart et al., 2018; Lovell et al., 2018). 6. Conclusions High-resolution, remotely sensed images from the combined terrestrial and submarine glacier forefields in the innermost part of Trygghamna are used to produce a composite geomorphological map and conceptual landsystem model of the fjord. The map distinctly demonstrates the contrasts between glacial geomorphological archives in the terrestrial and submarine environments. • Harriet- and Kjerulfbreen have exhibited surge-behavior, based particularly on the presence of terrestrial CSRs. It cannot be concluded with certainty if Protektorbreen has surged in the past as other factors could affect the preservation potential and formation of landforms. The lack of CSRs in the submarine environment could owe to burial by glacial sediments or overprinting of the recessional moraines. • Crevasse-squeeze ridges (CSRs) are only observed in the terrestrial archive, and recessional moraines are only formed in the submarine

archive. This suggests differences in dynamic behavior during the retreat between the inactive cold-based terrestrial terminus of the glacier compared to the active retreat of the warm-based submarine terminus of the same glacier. • The landform assemblage in Trygghamna does not correspond entirely to any of the published surge-type landsystem models. Therefore, a conceptual landsystem model of surge-type glaciers in Trygghamna is presented, emphasizing the contrast between the terrestrial and submarine environments. • This study highlights that reconstructing glacial dynamics and history should, where possible, integrate evidence from both terrestrial and submarine archives to enable holistic assessments of past and present glaciated areas. Declaration of Competing Interest None. Acknowledgements We acknowledge Lis Allaart, Filip Johansson, Åsa Wallin, Paul Vesland, Sarah Strand and Trude Hole for contribution, assistance and constructive discussion during the field work. Andy Hodson is thanked for team work and the zodiac during field work in 2015. Erik Schytt Holmlund is acknowledged for providing historical data from Trygghamna, Sara Mollie Cohen for assistance with data, logistics and 3D computer at UNIS and Wesley R. Farnsworth for helpful feedback. The fieldwork logistics and safety were co-funded by UNIS research allocation to Ó. Ingólfsson (2015, 2016) and Arctic Field Grant (2016) (grant number 201610), which is gratefully acknowledged. The Ministry of Foreign Affairs in Iceland is also thanked for financial support during the research. Depth data are reproduced according to permission no. 13/G706 by the Norwegian Hydrographic Service. Two anonymous reviewers are thanked for their constructive comments that improved the manuscript. References Allaart, L., Friis, N., Ingólfsson, Ó., Håkansson, Noormets, R., Farnsworth, W.R., Mertes, J., Schomacker, A., 2018. Drumlins in the Nordenskiöldbreen forefield, Svalbard. GFF 140 (2). Batchelor, C., Dowdeswell, J.A., Ottesen, D., 2017. Submarine glacial landforms. In: Micallef, A., Krastel, S., Savini, A. (Eds.), Submarine Geomorphology. Springer, Cham, Switzerland, pp. 207–234.


N. Aradóttir et al. / Geomorphology 344 (2019) 75–89

Benediktsson, Í.Ö., Möller, P., Ingólfsson, Ó., van der Meer, J.J.M., Kjær, K.H., Krüger, J., 2008. Instantaneous end moraine and sediment wedge formation during the 1890 glacier surge of Brúarjökull, Iceland. Quat. Sci. Rev. 27 (3–4), 209–234. https://doi. org/10.1016/j.quascirev.2007.10.007. Benediktsson, Í.Ö., Schomacker, A., Lokrantz, H., Ingólfsson, Ó., 2010. The 1890 surge end moraine at Eyjabakkajökull, Iceland: a re-assessment of a classic glaciotectonic locality. Quat. Sci. Rev. 29 (3), 484–506. Benediktsson, Í.Ö., Schomacker, A., Johnson, M.D., Geiger, A.J., Ingólfsson, Ó., Guðmundsdóttir, E.R., 2015. Architecture and structural evolution of an early Little Ice Age terminal moraine at the surge-type glacier Múlajökull, Iceland. J. Geophys. Res. Earth Surf. 120, 1895–1910. Benn, D.I., Evans, D.J.A., 2010. Glacier and Glaciations. Second edition. Hodder Education, London. Bennett, M.R., 2001. The morphology, structural evolution and significance of push moraines. Earth Sci. Rev. 53 (3–4), 197–236. Ben-Yehoshua, D., 2017. Crevasse-Squeeze Ridges in Trygghamna. Thesis, University of Iceland, Reykjavik, Svalbard. MSc. Boulton, G.S., 1968. Flow tills and related deposits on some Vestspitsbergen glaciers. J. Glaciol. 7 (51), 391–412. Boulton, G.S., 1976. The origin of glacially fluted surfaces-observations and theory. J. Glaciol. 76, 287–309. Boulton, G.S., 1986. Push-moraines and glacier-contact fans in marine and terrestrial environments. Sedimentology 33 (5), 677–698. Boulton, G.S., van de Meer, J.J.M., Harts, J.K., Beets, D.J., Ruegg, G.H.J., van der Wateren, F.M., Jarvis, J., 1996. Till and moraine emplacement in a deforming bed surge - an example from a marine environment. Quat. Sci. Rev. 15 (95), 961–987. 10.1016/0277-3791(95)00091-7. Boulton, G.S., van de Meer, J.J.M., Beets, D.J., Hart, J.K., Ruegg, G.H.J., 1999. The sedimentary and structural evolution of a recent push moraine complex: Holmstrømbreen, Spitsbergen. Quat. Sci. Rev. 18 (3), 339–371. 00068-7. Bouvier, V., Johnson, M. D., Påsse, T., 2015. Distribution, genesis and annual-origin of De Geer moraines in Sweden: insights revealed by LiDAR. GFF 1-15. doi:https://doi. org/10.1080/11035897.2015.1089933. Brynjólfsson, S., 2015. Dynamics and Glacial History of the Drangajökull Ice Cap, Northwest Iceland. Norwegian University of Science and Technology, Trondheim. (Doctoral dissertation, 233 pp). Brynjólfsson, S., Ingólfsson, Ó., Schomacker, A., 2012. Surge fingerprinting of cirque glaciers at the Tröllaskagi. Jökull 62, 153–168. Brynjólfsson, S., Schomacker, A., Ingólfsson, Ó., 2014. Geomorphology and the Little Ice Age extent of the Drangajökull ice cap, NW Iceland, with focus on its three surge-type outlets. Geomorphology 213, 292–304. Burton, D.J., Dowdeswell, J.A., Hogan, K.A., Noormets, R., 2016. Marginal fluctuations of a Svalbard surge-type tidewater glacier, Blomstrandbreen, since the Little Ice Age: a record of three surges. Arct. Antarct. Alp. Res. 48 (2), 411–426. AAAR0014-094. Chandler, B.M.P., Lovell, H., Boston, C.M., Lukas, S., Barr, I.D., Benediktsson, Í.Ö., Benn, D.I., Clark, C.D., Darvill, C.M., Evans, D.J.A., Ewertowski, M.W., Loibl, D., Margold, M., Otto, J.-C., Roberts, D.H., Stokes, C.R., Storrar, R.D., Stroeven, A.P., 2018. Glacial geomorphological mapping: a review of approaches and frameworks for best practice. Earth-Sci. Rev. 185, 806–846. de Geer, G., 1910. Guide de l’excursion au Spitsberg. Excursion A1. XI International Geological Congress, Stockholm, Sweden. Dickson, R.R., Osborn, T.J., Hurrell, J.W., Meincke, J., Blindheim, J., Adlandsvik, B., Vinje, T., Alekseev, G., Maslowski, W., 2000. The Arctic Ocean response to the North Atlantic Oscillation. J. Clim. 13 (15), 2671–2696. 013b2671:TAORTTN2.0.CO;2. Dowdeswell, J.A., Hodgkins, R., Nuttall, A., Hagen, J.O., Hamilton, G.S., 1995. Mass balance change as a control on the frequency and occurrence of glacier surges in Svalbard, Norwegian High Arctic. Geophys. Res. Lett. 22 (21), 2909–2912. 10.1029/95GL02821. Dowdeswell, J.A., Canals, M., Jakobsson, M., Todd, B.J., Dowdeswell, E.K., Hogan, K.A., 2016. Atlas of submarine glacial landforms: modern. Quaternary and Ancient. Geological Society, London Memoirs, No. 46. Dyke, A.S., 1993. Landscapes of cold-centred Late Wisconsinan ice caps, Arctic Canada. Prog. Phys. Geogr. 17 (2), 223–247. Evans, D.J.A. (Ed.), 2003. Glacial Landsystems. Arnold, London, p. 532. Evans, D.J.A., Rea, B.R., 1999. Geomorphology and sedimentology of surging glaciers: a land-systems approach. Ann. Glaciol. 28 (1), 75–82. 172756499781821823. Evans, D.J.A., Rea, B.R., 2003. Surging glacier landsystem. In: Evans, D.J.A. (Ed.), Glacial Landsystems. Arnold, London, pp. 259–288 Evans, D.J.A., Twigg, D.R., 2002. The active temperate glacial landsystem: a model based on Breiðarmerkurjökull and Fjallsjökull, Iceland. Quat. Sci. Rev. 21, 2143–2177. Eyles, N., Rogerson, R.J., 1978. A framework for the investigation of medial moraine formation: Austerdalsbreen, Norway, and Berendon Glacier, British Columbia, Canada. J. Glaciol. 20 (82), 99–113. Farnsworth, W.R., Ingólfsson, Ó., Retelle, M., Schomacker, A., 2016. Over 400 previously undocumented Svalbard surge-type glaciers identified. Geomorphology 264, 52–60. Farnsworth, W.R., Ingólfsson, Ó., Noormets, R., Allaart, L., Alexanderson, H., Henriksen, M., Schomacker, A., 2017. Dynamic Holocene glacial history of St. Johnsfjorden, Svalbard. Boreas 46 (3), 585–603.

Flink, A.E., Noormets, R., Kirchner, N., Benn, D.I., Luckman, A., Lovell, H., 2015. The evolution of a submarine landform record following recent and multiple surges of Tunabreen glacier, Svalbard. Quat. Sci. Rev. 108, 37–50. quascirev.2014.11.006. Flink, A.E., Noormets, R., Fransner, O., Hogan, K.A., O'Regan, M., Jakobsson, M., 2017. Past ice flow in Wahlenbergfjorden and its implications for late Quaternary ice sheet dynamics in northeastern Svalbard. Quat. Sci. Rev. 163, 162–179. 10.1016/j.quascirev.2017.03.021. Forwick, M., 2005. Sedimentary processes and paleoenvironments in Spitsbergen fjords. University of Tromsø. (Doctoral dissertation, 173 pp). Forwick, M., Vorren, T.O., 2010. Stratigraphy and deglaciation of the Isfjorden area, Spitsbergen. Nor. J. Geol. 90, 163–179. Fransner, O., Noormets, R., Flink, A.E., Hogan, K.A., O'Regan, M., Jakobsson, M., 2017. Glacial landforms and their implications for glacier dynamics in Rijpfjorden and Duvefjorden, northern Nordaustlandet, Svalbard. J. Quat. Sci. 32, 437–455. https:// Glasser, N.F., Hambrey, M.J., 2003. Ice-marginal terrestrial landsystems: Svalbard polythermal glaciers. In: Evans, D.J.A. (Ed.), Glacial Landsystems. Arnold, London, pp. 65–88. Glasser, N.F., Hambrey, M.J., Crawford, K.R., Bennett, M.R., Huddart, D., 1998. The structural glaciology of Kongsvegn, Svalbard, and its role in landform genesis. J. Glaciol. 44 (146), 136–148. Greenwood, S.L., Clark, C.F., Hughes, A.L.C., 2007. Fomalising an inversion methodology for reconstructing ice-sheet retreat patterns from meltwater channels: application to the British Ice Sheet. J. Quat. Sci. 22. 637-345. Hagen, J.O., Liestøl, O., Roland, E., Jørgensen, T., 1993. Glacier Atlas of Svalbard and Jan Mayen. Norsk Polarinstitutt Meddeleser 129. Norwegian Polar Institute, Oslo. Hagen, J.O., Kohler, J., Melvold, K., Winther, J.-G., 2003. Glaciers in Svalbard: mass balance, runoff and freshwater flux. Polar Res. 22 (2), 145–159. j.1751-8369.2003.tb00104.x. Hamberg, 1905. Karte der Baie recherché und van Keulen Bay auf Spitzbergen wahrend der von A. G. Nathorst geleiteten Schwedischen Polarexpedition 1898 photogrammetrisch aufgenommen und konstruiert von axel Hamber. Scale 1:100 00. –A. Hamberg, Astronomische, photogrammetrisch und erdmagnetische Arbeiten der von A. G. nathorst geleiteten schwedischen Polarexpedition 1898. Uppsala 1905. [Pl. 1] (vet-ak. Handl. N. F. 39. No. 6.) Hambrey, M.J., Huddart, D., Bennett, M.R., 1997. Genesis of ‘hummocky moraines’ by thrusting in glacier ice: evidence from Svalbard and Britain. J. Geol. Soc. 154, 623–632. Hambrey, M.J., Bennett, M.R., Dowdeswell, J.A., Glasser, N.F., Huddart, D., 1999. Debris entrainment and transfer in polythermal valley glaciers. J. Glaciol. 45 (149), 69–86. Hanssen-Bauer, M., Kristensen, S., Steffensen, E.L., 1990. The climate of Spitsbergen. Klima, Den Norske Meteorologiske Institutt Rapport 39, 1–40. Holtedahl, O., 1913. Geologische Karte tiber das Gebiet ostlich des Vorlandsundes auf Spitsbergen. Masst. 1 (200), 000. Isachsen, G., 1915. Ocean Glacial arctique. Spitsberg. [Partie Nord-Ouest.] Carte dressee sous la direction du captiaine Gunnar Isachsen par l'expedition Norvegienne. 1909-1910. Echelle 1:200 000. – Aiso in: Gunnar Isachsen. Travaux topographiques de l'expedition Iscahsen. 1909-1910. Kristiania 1915. [Annexure.] (Vid.-selsk. Skr. I. 1915. No. 7). Ives, L.R.W., Iverson, N.R., 2019. Genesis of glacial flutes inferred from observations at Múlajökull, Iceland. Geology 47 (5), 387–390. Jiskoot, H., Boyle, P., Murray, T., 1998. The incidence of glacier surging in Svalbard: evidence from multivariate statistics. Comput. Geosci. 24 (4), 387–399. 10.1016/S0098-3004(98)00033-8. Jónsson, S.A., Schomacker, A., Benediktsson, Í.Ö., Ingólfsson, Ó., Johnson, M.D., 2014. The drumlin field and the geomorphology of the Múlajökull surge-type glacier, Central Iceland. Geomorphology 207, 213–220. geomorph.2013.11.007. Kamb, B., Raymond, C.F., Harrison, W.D., Engelhardt, H., Echelmeyer, K.A., Humphrey, N., Brugman, M.M., Pfeffer, T., 1985. Glacier Surge Mechanism: 1982-1983 Surge of Variegated Glacier, Alaska. Science 227(4686), 1982–1983. science.227.4686.469. Kjær, K.H., Krüger, J., 2001. The final phase of dead-ice moraine development: processes and sediment architecture, Kötlujökull, Iceland. Sedimentology 48, 935–952. Kjær, K.H., Larsen, E., van der Meer, J., Ingólfsson, Ó., Krüger, J., Benediktsson, Í.Ö., Knudesen, C.G., Schomacker, A., 2006. Subglacial decoupling at the sediment / bedrock interface: a new mechanism for rapid flowing ice. Quat. Sci. Rev. 25, 2704–2712. Koenig, M. (submitter); Koenig, M.; Nuth, C. (analyst(s)), 2010. GLIMS Glacier Database. Boulder, CO. National Snow and Ice Data Center. doi: N5V98602. Kristensen, L., Benn, D.I., Hormes, A., Ottesen, D., 2009. Mud aprons in front of Svalbard surge moraines: evidence of subglacial deforming layers or proglacial glaciotectonics? Geomorphology 111 (3–4), 206–221. geomorph.2009.04.022. Krüger, J., Kjær, K.H., 1999. A data chart for field description and genetic interpretation of glacial diamicts and associated sediments – with examples from Greenland, Iceland and Denmark. Boreas 28 (3), 386–402. tb00228.x. Lefauconnier, B., Hagen, J.O., 1991. Surging and calving glaciers in Eastern Svalbard. Norsk Polarinstitutt Meddeleser 116. Norwegian Polar Institute, Oslo. Liestøl, O., 1969. Glacier surges in west Spitsbergen. Can. J. Earth Sci. 6 (4), 895–897.

N. Aradóttir et al. / Geomorphology 344 (2019) 75–89 Liestøl, O., 1988. The glaciers in Kongsfjorde area, Spitsbergen. Nor. Geogr. Tidsskr. 42, 231–238. Liljequist, G.H., 1993. High Latitudes: A History of Swedish Polar Travels and Research. The Swedish Polar Research Secretariat. [Polarforskningssekretariatet], Stockholm. Lindén, M., Möller, P., 2005. Marginal formation of De Geer moraines and their implications to the dynamics of grounding-line recession. J. Quat. Sci. 20 (2), 113–133. Lønne, I., 2016. A new concept for glacial geological investigations of surges, based on High-Arctic examples (Svalbard). Quat. Sci. Rev. 132, 74–100. 10.1016/j.quascirev.2015.11.009. Lovell, H., Boston, C.M., 2017. Glaciotectonic composite ridge system and surge-type glaciers: an updated correlation based on Svalbard, Norway. Arktos 3 (2), 1–16. https:// Lovell, H., Fleming, E.J., Benn, D.I., Hubbard, B., Lukas, S., Rea, B.R., Noormets, R., Flink, A.E., 2015. Debris entrainment and landform genesis during tidewater glacier surges. J. Geophys. Res. Earth Surf. 120 (8), 1574–1595. 2015JF003509. Lovell, H., Benn, D.I., Lukas, S., Ottesen, D., Luckman, A., Hardiman, M., Barr, I.D., Boston, C.M., Sevestre, H., 2018. Multiple Late Holocene surges of High-Arctic tidewater glacier systems in Svalbard. Quat. Sci. Rev. 201, 162–186. quascirev.2018.10.024. Maclachlan, S.E., Howe, J.A., Vardy, M.E., 2010. Morphodynamic evolution of Kongsfjorden-Krossfjorden, Svalbard, during the Late Weichselian and Holocene. Geol. Soc. Lond., Spec. Publ. 344 (1), 195–205. Meier, M.F., Post, A., 1969. What are glacier surges? Can. J. Earth Sci. 6 (4), 807–817. Murray, T., Strozzi, T., Luckman, A., Jiskoot, H., Christakos, P., 2003. Is there a single surge mechanism? Contrasts in dynamics between glacier surges in Svalbard and other regions. Journal of Geophysical Research 108(B5, 2237), 1–15. doi: 10.1029/2002JB001906. Nordenskiöld, G., 1892. Redogörelse för den svenska expeditionen till Spetsbergen 1890: med sex taflor och en karta (Stockholm). Norwegian Polar Institute, 2003. The Place Names of Svalbard. Rapportserie Nr. 122. Norwegian Polar Institute, Tromsø. Nuth, C., Kohler, J., König, M., von Deschwanden, A., Hagen, J.O., Kääb, A., Moholdt, G., Pettersson, R., 2013. Decadal changes from a multi-temporal glacier inventory of Svalbard. Cryosphere 7, 1603–1621. Ó Cofaigh, C., Lemman, D.S., Evans, D.J.A., Bednarski, J., 1999. Glacial landform-sediment assemblages in the Canadian High Arctic and their implications for late Quaternary glaciations. Ann. Glaciol. 28, 195–201. Ohta, Y., Hjelle, A., Andresen, A., Dallmann, W.K., Salvigsen, O., 1992. Geological Map Svalbard 1:100,000, sheet B9G Isfjorden. Norsk Polarinstitutt Temakart Nr. 16. Ottesen, D., Dowdeswell, J.A., 2006. Assemblages of submarine landforms produced by tidewater glaciers in Svalbard, J. Geophys. Res. 111, 1–16. doi: 10.1029/2005JF000330. Ottesen, D., Dowdeswell, J.A., Benn, D.I., Kristensen, L., Christiansen, H.H., Christensen, O., Hansen, L., Lebesbye, E., Forwick, M., Vorren, T.O., 2008. Submarine landforms characteristic of glacier surges in two Spitsbergen fjords. Quat. Sci. Rev. 27 (15–16), 1583–1599. Ottesen, D., Dowdeswell, J.A., Bellec, V., Bjarnadóttir, L., 2017. The geomorphic imprint of glacier surges into open-marine waters: examples from eastern Svalbard. Mar. Geol. 392, 1–29.


Plassen, L., Vorren, T.O., Forwick, M., 2004. Integrated acoustic and coring investigation of glacigenic deposits in Spitsbergen fjords. Norsk Polarinstitutt 23 (1), 89–110. https:// Raymond, C.F., 1987. How do glaciers surge? A review. J. Geophys. Res. 92 (B9), 9121–9134. Rea, B.R., Evans, D.J.A., 2011. An assessment of surge-induced crevassing and the formation of crevasse squeeze ridges. J. Geophys. Res. Earth Surf. 116, 1–17. https://doi. org/10.1029/2011JF001970. Salvigsen, O., Elgersma, A., Hjort, C., Legerlund, E., Liestøl, O., Svensson, N., 1990. Glacial history and shoreline displacement on Bohemanflya, Spitsbergen, Svalbard. Polar Res. 8, 261–273. Schomacker, A., Kjær, K.H., 2007. Origin and de-icing of multiple generations of ice-cored moraines at Brúarjökull, Iceland. Boreas 36 (4), 411–425. 03009480701213554. Schomacker, A., Kjær, K.H., 2008. Quantification of dead-ice melting in ice-cored moraines at Holmströmbreen, Svalbard. Boreas 37 (2), 211–225. j.1502-3885.2007.00014.x. Schomacker, A., Benediktsson, Í.Ö., Ingólfsson, Ó., 2014. The Eyjabakkajökull glacial landsystem, Iceland: geomorphic impact of multiple surges. Geomorphology 218, 98–107. Serreze, M.C., Maslanik, J.A., Scambos, T.A., Fetterer, F., Stroeve, J., Knowles, K., Fowler, C., Drobot, S., Barry, R.G., Haran, T.M., 2003. A record minimum arctic sea ice extent and area in 2002. Geophys. J. 30 (3). Sevestre, H., Benn, D.I., 2015. Climatic and geometric controls on the global distribution of surge-type glaciers: implications for a unifying model of surging. J. Glaciol. 61 (228), 646–662. Sevestre, H., Benn, D.I., Hulton, N.R.J., Bælum, K., 2015. Thermal structure of Svalbard glaciers and implications for thermal switch models of glacier surging. J. Geophys. Res. Earth Surf. 120, 2220–2236. Sharp, M., 1985. “Crevasse-fill” ridges-a landform type characteristic of surging glaciers? Geografisker Annaler, Series A. Phys. Geogr. 67 (3–4), 213–220. 10.1080/04353676.1985.11880147. Sharp, M., 1988. Surging glaciers: geomorphic effects. Prog. Phys. Geogr. 12 (4), 533–559. Solheim, A., 1991. The depositional environment of surging sub-polar tidewater glaciers, vol. 194. Norsk polarinstitutt skrifter, (Norwegian polarinstitute), pp. 1-97. Streuff, K., Ó Cofhaig, C., Noormets, R., Lloyd, J., 2018. Submarine landform assemblages and sedimentary processes in front of Spitsbergen tidewater glaciers. Mar. Geol. 402, 209–227. Sund, M., Eiken, T., Rolstad, C.D., 2011. Velocity structure, front position changes and calving of the tidewater glacier Kronebreen, Svalbard. Cryosphere 5, 41–73. https://doi. org/10.5194/tcd-5-41-2011. Svendsen, J.I., Mangerud, J., 1997. Holocene glacial and climatic variations on Spitsbergen, Svalbard. The Holocene 7 (1), 45–57. Wallin, Å.C., 2016. Investigating methods for paleo surge-type glaciers of highly dynamical ice flows in Trygghamna, west Spitsbergen. University of Stockholm, MSc. Thesis Werner, A., 1993. Holocene moraine chronology, Spitsbergen, Svalbard: lichenometric evidence for multiple Neoglacial advances in the Arctic. The Holocene 2, 128–137. Yde, J., Paasche, Ø., 2010. Reconstructing climate change: not all glaciers suitable. Eos 91 (21), 189–196.