Master’s thesis
Physical Geography and Quaternary Geology, 60 Credits
Investigating methods for identifying paleo surge-type glaciers or highly dynamical ice flows in Trygghamna, west
Spitsbergen
Åsa Cecilia Wallin
NKA 150
2016
Preface
This Master’s thesis is Åsa Cecilia Wallin’s degree project in Physical Geography and Quaternary Geology at the Department of Physical Geography, Stockholm University. The Master’s thesis comprises 60 credits (two terms of full-time studies).
Cooperation with UNIS, University Centre in Svalbard.
Supervisors have been Jan Risberg at the Department of Physical Geography, Stockholm University, and Andy Hodson at the University of Sheffield and Harold Lovell at University of Portsmouth.
Examiner has been Per Holmlund at the Department of Physical Geography, Stockholm University.
The author is responsible for the contents of this thesis.
Stockholm, 20 June 2016
Steffen Holzkämper Director of studies
Surge-‐type glaciers exhibit a cyclic behaviour with an ice mass increase in the reservoir area during the inactive, quiescent phase and a rapid transportation of ice during the active, surge phase. In or-‐
der to interpret the effects of climate change correctly it is important to distinguish between advanc-‐
es of surge-‐type glaciers and those of ‘normal’ glaciers, caused by climatic fluctuations. This is partic-‐
ularly important for the Arctic, which is predicted to experience the highest increase in temperature on the planet.
The dynamic and mechanism of surge-‐type glaciers can be used for understanding both modern and past ice sheet dynamic instabilities, threshold behavior and contribution to sea level rise. They are also suggested to be analogues for land-‐terminating paleo ice streams and surging ice sheet lobes, which makes them highly valuable as research objects. Though in order to understand their behavior, it is important to be able to identify them. In literature the number for surge-‐type glaciers on Svalbard varies between 13 and 90 %, thus it is important to work out methods, other than physical observations, for identification. Aerial photographs are a powerful tool for the identifi-‐
cation and mapping of landforms and ice structures and for reconstructing glacier distribution.
Though for mapping of ice facies and structures within the glaciers, which provide important infor-‐
mation regarding the glaciers past dynamics, and is an important part of the identification, fieldwork is essential. Fieldwork is also much needed in order to interpret the landforms and their genesis cor-‐
rectly. It is suggested that the use of multiple methods for identifying surges, or highly dynamical ice flows, will improve the result and the reliability will increase.
In this thesis two different methods, structural glaciology and glacial geomorphology, have been used for interpreting the past behaviour of the four glaciers in Trygghamna, west Spitsbergen.
Structural glaciology provides a way to determine the dynamic of surge-‐type glaciers for both the quiescent-‐ and surge phase. By using glacial geomorphology it is possible to reconstruct former ex-‐
tent and thickness of small valley glaciers. The glaciers of Trygghamna exhibit evidence of a dynamic past, of which Harrietbreen and Kjerulfbreen bear enough evidence to be considered as surge-‐type glaciers. The lack of landforms on the forelands of Kiærbreen and Protektorbreen makes the inter-‐
pretation difficult, though the ice facies and structures within the ice cave on Protektor-‐breen pro-‐
vide additional information and thus this glacier should at least be classified as a highly dynamical ice flow. Thus it is considered to not be suitable to only use one data type to determine whether a glaci-‐
er is of surge-‐type or not. By use multiple methods the result will improve and the reliability will in-‐
crease. The two methods used for this thesis are considered to be the least amount of methods needed in order to establish whether a glacier is of surge-‐type or not.
I would like to thank my supervisors whose help has been extremely valuable for this thesis. Harold Lovell, for the invaluable help of understanding and interpreting the ice facies and structures. It has been a fantastic learning curve about something that I genuinely find fascinating, but had absolutely no knowledge about at the start. Without his feedback I would not only have grey hair, but no hair what so ever. Jan Risberg for keeping me on track and not drift off and providing his expertise regard-‐
ing writing and language usage. Andy Hodson, without whom I would never have come to Tryggham-‐
na and fallen in love with the place. Also a huge thanks to him and Ólafur Ingólfsson making it possi-‐
ble for me to spend two weeks in Trygghamna for fieldwork. They provided a fantastic experience, just four master students, one field assistant and two dogs on our own in the land of polar bears, ending this adventurous trip with a minor hurricane ripping the tents to pieces. Absolutely loved it and wouldn’t like to have it any other way! Of course my eminent field assistant Trude Hohle de-‐
serves a big thank you and an even greater hug. Without her I would not have dared to camp far out in the wilderness, but together we can manage anything! Polar bears, hurricanes, mad people -‐ noth-‐
ing is too difficult for us, my dear. My fellow master students Nina Aradottir, Daniel Ben Yeoshua and Filip Johansson for long and fruitful discussions. I wish you all a very good luck with your thesis’s!
Heïdi Sevestre, for believing in the project and eagerly encouraging me to try out this project. The De Geer Foundation for financial support, making it possible for me to get the data needed for the the-‐
sis. My loving family who has had to accept that their youngest member has become bitten by the polar bug and decided to live up in the high Arctic. You might not really understand me, but thank you for accepting that this is what I love and want to continue to do. Love you to bits!
1 INTRODUCTION 1
2 STUDY AREA 3
2.1 THE PHYSICAL GEOGRAPHY OF SVALBARD 4
2.2 CLIMATIC CONDITIONS AND SENSITIVITY OF THE SVALBARD ARCHIPELAGO 5
2.3 TRYGGHAMNA 6
2.4 THE GLACIERS OF TRYGGHAMNA, THEIR GROWTH AND RETREAT 8
2.4.1 PROTEKTORBREEN 10
2.4.2 HARRIETBREEN 10
2.4.3 KJERULFBREEN 11
2.4.4 KIÆRBREEN 11
3 METHODS 12
3.1 DATA SOURCES 12
3.1.1 PUBLISHED DATA 12
3.1.2 AERIAL PHOTOGRAPHS 12
3.1.3 GLACIAL GEOMORPHOLOGICAL AND STRUCTURAL GLACIOLOGICAL MAPPING 12
4 DEFINITION AND CHARACTERISTICS FOR GLACIERS AND SURGE-‐TYPE GLACIERS, IN GENERAL
AND SVALBARD IN SPECIFIC 13
4.1 GLACIERS 14
4.1.1 AREAS ON A GLACIER 14
4.1.2 THERMAL STATE OF GLACIERS 15
4.1.3 GLACIERS ON SVALBARD 15
4.2 SURGE-‐TYPE GLACIERS 16
4.2.1 SURGE CYCLE 16
4.2.2 MECHANISMS FOR SURGE-‐TYPE GLACIERS 18
5 GLACIAL GEOMORPHOLOGY – A TOOL FOR IDENTIFYING SURGES 19
5.1 DESCRIPTION OF LANDFORMS INDICATIVE FOR SURGE-‐TYPE GLACIERS AND IMPORTANT FOR RECONSTRUCTING
PAST EXTENT OF THE GLACIERS IN TRYGGHAMNA 20
5.1.1 LANDFORMS FOR RECONSTRUCTING PAST GLACIER EXTENTS, THICKNESSES AND FLOW DIRECTIONS 20
5.1.2 LANDFORMS INDICATIVE OF SURGE-‐TYPE GLACIERS 21
6 STRUCTURAL GLACIOLOGY 24
6.1 ICE STRUCTURES -‐ WITNESSES OF A GLACIERS PAST BEHAVIOUR 25
6.1.1 S0 -‐ PRIMARY STRATIFICATION 27
6.1.2 S1 -‐ LONGITUDINAL FOLIATION 27
6.1.3 S2 -‐ ARCUATE FRACTURE TRACES 27
6.1.4 S3 -‐ FRACTURE TRACES 28
6.1.5 S4 -‐ OPEN FRACTURES 28
6.2 ICE FACIES 28
7 RESULT 30
7.1 STRUCTURAL GLACIOLOGY 31
7.1.1 ICE STRUCTURES ON THE SURFACE OF THE GLACIERS IN TRYGGHAMNA AND ICE FLOW DIRECTIONS 31 7.1.2 ICE STRUCTURES AND ICE FACIES WITHIN THE ICE CAVE ON PROTEKTORBREEN 33 7.2 LANDFORMS ON THE FORELANDS AND DEBRIS FEATURES ON THE GLACIER SURFACE 36
8 DISCUSSION 38
8.1 PROTEKTORBREEN 39
8.1.1 STRUCTURAL GLACIOLOGY 39
8.1.2 RIDGES ON THE FORELAND AND ON THE GLACIER INDICATORS OF PAST ICE DYNAMICS 44
8.1.3 SUMMARY PROTEKTORBREEN 44
8.2 HARRIETBREEN 45
8.2.1 ICE STRUCTURES ON THE GLACIER SURFACE INDICATING A HIGHLY DYNAMIC ICE FLOW 45
8.2.2 GLACIAL GEOMORPHOLOGICAL EVIDENCE ON THE FORELAND 46
8.2.3 SUMMARY HARRIETBREEN 47
8.3 KJERULFBREEN 47
8.3.1 A GLACIER SURFACE WITH A HIGH AMOUNT OF FRACTURE TRACES AND DEBRIS FEATURES 47
8.3.2 THE FORELAND AND ITS LANDFORMS 48
8.3.3 SUMMARY KJERULFBREEN 50
8.4 KIÆRBREEN 51
8.4.1 STRUCTURAL GLACIOLOGY PROVIDES MOST INFORMATION REGARDING PAST ICE FLOW BEHAVIOUR 51
8.4.2 THE ABSENCE OF GLACIAL GEOMORPHOLOGICAL EVIDENCE 51
8.4.3 SUMMARY KIÆRBREEN 52
8.5 COMPARING THE GLACIERS IN TRYGGHAMNA – SURGE-‐TYPE, HIGHLY DYNAMIC OR NONE OF THE ABOVE? 52
9 CONCLUSION 54
10 REFERENCES 55
11 APPENDIX 57
Though some glaciers accumulate intensely in the upper area and suddenly advance several meters per day, so called surge-‐type glaciers. For interpreting the effects of climate change it is important to distinguish between these types of glaciers and advances of ‘normal’ glaciers caused by climatic fluc-‐
tuations (Sharp, 1988). This is particularly important for the Arctic, which is predicted to experience the highest increase in temperature on the planet. It is also of importance to investigate the dynamic and mechanism of surge-‐type glaciers in order to understand the mechanisms of both modern and past ice sheet dynamic instabilities, threshold behavior and contribution to sea level rise (Ingólfsson et al., 2016). Surge-‐type glaciers are also suggested to be analogues for land-‐terminating paleo-‐ice streams and surging ice sheet lobes (Ingólfsson et al., 2016), which makes them even more valuable as research objects. Though in order to understand their behavior, it is important to be able to identi-‐
fy them. Several processes occur during a surge, there is not just one process unique to surges but certain combinations of processes, related to the shift between fast and slow ice flow. For identifying paleosurges it is the assemblage of landforms, sediments (Sharp, 1985) and ice structures (Lovell et al., 2015b) indicative of processes during a glacial cycle that provides the information needed.
In literature the number for surge-‐type glaciers on Svalbard varies between 13% and 90%
(Hagen & Liestøl, 1993; Jiskoot et al., 2000; Bennet & Glasser, 2009). This stress the importance of finding methods for identifying surge-‐type glaciers that has never been observed or documented to surge. Svalbard glaciers have experienced a rapid recession and are at present much less dynamic than in the past and are thus considered not be able to build up to surges as frequent as in the past.
Due to the retreat, most of the small valley glaciers on Svalbard have shifted from polythermal to cold based and are now frozen to their beds (Lovell et al., 2015a). Ice structures record a time when the glaciers were thermally different and subject to intense crevassing. It is important to reconstruct the timing and characteristics as it provides a link between glacier thermal regime and climate cycles, which affect flow dynamics and thermally controlled surge behavior. The documentation of distribu-‐
tion and thicknesses of glaciers is valuable information for interpreting what have happened with the glaciers on Svalbard since the Little Ice Age (LIA). Though, many surge-‐type glaciers are located in remote areas where direct measurements are difficult to conduct. This makes it easy to miss a surge, or it might already be in the active phase of a surge before it is detected. By the use of aerial photo-‐
graphs glacier dynamics in these areas can be studied (Murray et al., 2003). Aerial photography is also a powerful tool for the identification and mapping of landforms and ice structures and for recon-‐
struction of glacier distribution. By using glacial geomorphology it is possible to reconstruct former extent and thickness of small valley glaciers (Lovell et al., 2015a). Though in order to understand the glacial landscape it is also important to understand the dynamics of the ice that covered it (Evans &
Rea, 1999). By documenting and analysing crevasse patterns of glacier surfaces this can be achieved.
The spatial resolution of aerial photographs can provide information that cannot be obtained by sat-‐
ellite imagery, such as glacier surface texture, glacier structures and interactions with the surround-‐
ing environment. They are particularly useful for mapping in remote or inaccessible areas (Hubbard and Glasser, 2005). Structural glaciology provides a way to determine the dynamic of surge-‐type glac-‐
iers for both the quiescent-‐ and active phase (Hambrey & Dowdeswell, 1997).
The aim of this thesis is to try out methods for identifying glaciers that have undergone surg-‐
es, or at least a highly dynamical ice flows. For this purpose, the glaciers in Trygghamna works per-‐
fectly since there are no historical records found, except for maps made by De Geer (1910) and Norsk Polarinstitutt (1936). These maps demonstrate a highly glaciated area where three of the four glaci-‐
ers terminated in the fjord. Their present surfaces bear witness of a dynamic ice flow and some of the forelands withhold landforms that support this, though not all. The methods used for the project are structural glaciology, both on the glacier surface and within the ice, together with glacial geo-‐
morphology, in particular the landforms associated with surges but also those indicating the extent and distribution of the glaciers. The hypothesis is that all four glaciers are of surge-‐type and that surges have caused the extensive loss of ice mass the past 100 years.
2 Study area
“Surrounded by glaciers and ice and pointed mountains. It is autumn.
The crowds of birds at Alkhornet are gone; red and green snow.”
Nathorst, 1882 (in Liljequist, 1993)
This part describes the geographical setting of Svalbard and in more detail the specific study area, Trygghamna, Spitsbergen.
Norsk Polarinstitutt, 2015
2.1 The physical geography of Svalbard
The Svalbard archipelago (77-‐80°N, 10-‐35°E) is located at the boundary of the Norwegian Sea, the Barents Sea and the Arctic Ocean (fig 1a, Humlum et al., 2005). The archipelago consists of several islands of which the main island Spitsbergen is the largest, covering 62% of the total land area of Svalbard (60 667 km2, fig 1b)). The first mentioning of Svalbard (meaning the ‘cool coast’) is in the Icelandic Annals of 1194, though Svalbard might have been discovered much earlier by the Norse Vikings (Hoel, 1942; Harland et al., 1997; Dallman et al., 2015). It is a highly glaciated, approximately 2100 glaciers (Worsley et al., 1986), and mountainous archipelago, with a coastline dominated by fjords and cliffs and it is suggested to have remained the same ever since the Pleistocene ice ages. A glacial minimum occurred after the last glaciation, approximately 7-‐8000 cal yr BP. Since then, an-‐
other two glacial maximum has occurred, the first about 2500 cal yr BP and the second during the Little Ice Age (LIA), from AD 1870-‐1920. The area covered by glaciers has been reduced from 61 % to 59 % during the last 40-‐ to 50 years (Dallman et al., 2015).
Svalbard is located in the northwestern corner of the Barents Sea Shelf, about 650 km north of Norway (Harland et al., 1997), and was uplifted by Late Mesozoic and Cenozoic crustal move-‐
ments. The Barents Sea Shelf between Svalbard and Fennoscandia is a platform area where Precam-‐
brian crust is mainly buried under a thick pile of Late Palaeozoic to Neogene sedimentary rocks (Dallman et al., 2015). The western side of Spitsbergen contains an extension of the Caledonian Orogeny, which was a mountain building event that occurred during the Ordovician and Silurian (~480-‐390 ma). By the end of the Silurian most of Svalbard’s rocks were metamorphosed and/or folded and faulted and younger sediments were deposited on top of the deformed succession, and later became bedrock (Dallman et al., 2015). The latest main deformation event in Svalbard, prelimi-‐
nary dated to the Eocene, is associated with the strike-‐slip displacement between Svalbard and Greenland during the Cenozoic (Harland et al., 1997; Dallman et al., 2015).
Figure 1 a) Svalbard is located about 650 km north of Norway, at the boundary of the Norwegian Sea, the Barents Sea
a b
2.2 Climatic conditions and sensitivity of the Svalbard archipelago
Svalbard is a well-‐studied part of the Arctic, with relatively mild climatic conditions due to the Nor-‐
wegian Current, which is a branch of the Gulf Stream, bringing up warmer water, creating the north-‐
ernmost area of open water in the Arctic winter. This result in temperatures above freezing also dur-‐
ing winter, although snow can fall during the short summer (Worsley et al., 1986; Hagen & Liestøl, 1993; Harland et al., 1997; Bennett & Glasser, 2009). The climate on Svalbard is variable, the west coast of Spitsbergen has a mean annual air temperature (MAAT) of -‐6°C together with a low rate of precipitation, 400-‐600 mm/yr, though the precipitation increase on the glaciers due to the orograph-‐
ic effect (fig 2). Further inland it is slightly colder and more continental (Hagen & Liestøl, 1993; Ham-‐
brey & Dowdeswell, 1997; Glasser et al 1998; Bennett & Glasser, 2009) and according to the Köppen-‐
Geiger classification – it is classified as ET, a polar tundra climate with cold winters and summers with a temperature less than +10°C (Kottek et al., 2006). These low temperatures make the ground per-‐
manently, or more correctly perennially, frozen, so called permafrost. The average depth of the per-‐
mafrost on Svalbard is ~300 m, (Worsley et al., 1986; Harland et al., 1997) ranging from <100 m at the coasts to >500 m in the highlands (Humlum et al. 2003).
Svalbard has a unique climatic sensitivity, due to its position at the northern extremity of relatively warm ocean currents and atmospheric depression tracks, and meteorological data has been well documented since 1911 (Humlum et al., 2003). The climatic sensitivity of Svalbard is prob-‐
ably caused by i) the archipelago is located in the main pathway of air masses into the Arctic Basin, ii) Svalbard is also located near the confluence of air masses and ocean currents with very different temperature characteristics and iii) it is also enhanced by the sea ice extent coupled with both at-‐
mospheric and oceanic circulation (Humlum et al., 2003). This makes Svalbard a key site for under-‐
standing the implications of climate change in the Northern Hemisphere (NH), especially since it is predicted an amplified climatic response in the Arctic (Humlum et al., 2005).
An interesting event occurred around 1920 as an abrupt warming, when the mean annual temperature changed from -‐9 to -‐4°C within five years (Dowdeswell et al., 1995), which is among the highest increase in surface temperature documented anywhere during this time period. This is re-‐
garded as the end of the Little Ice Age (LIA) in Svalbard. Between 1957 and 1968, the temperature decreased with about 5°C, with a more gradual increase in temperature until the end of the 20th cen-‐
tury (Humlum et al., 2003 & 2005).
Figure 2 Clouds building up over Protektorbreen due to the orographic effect.
2.3 Trygghamna
“A Safe Haven, well protected bay from most winds”
Orvin, 1991
Trygghamna is situated on the west coast of Spitsbergen, 41 km northwest of Longyearbyen at 78°14’N, 13°49’E (fig 3a). It is a small fjord, 5 km long and 2 km wide (fig 3a & b), on the north side of the outer part of Isfjorden (Hoel, 1942). The precipitation is a bit less than the average for the west coast, 373 mm (Hambrey & Dowdeswell, 1997; Glasser et al 1998). Though, according to Hagen &
Liestøl (1993), Trygghamna is situated where precipitation is suggested to 600 mm/yr, which can be caused by the orographic effect (fig 2). The glaciers here are either polythermal or cold-‐based, (Glasser et al 1998: Hambrey & Dowdeswell, 1997).
This fjord was a well-‐protected anchorage for ships during storms, formerly used by whalers.
Its first mention was by Dutch whalers referring to it as Behouden Haven in 1612 (Hoel, 1942; Ma-‐
thisen, 1956) and it has after that been translated into English and referred to as Safe Harbour, Safe Haven or Safe Bay (The Svalbard Commissioners, 1927). After the Svalbard Treaty in 1920, the place names in Svalbard changed from mainly English to Norwegian and Trygghamna became the official name of the fjord (The Svalbard Commissioners, 1927).
Whaling was a big industry in Svalbard from the 1600’s, and there are records from 1612 that the Dutch were in Trygghamna. The following years Englishmen, Dutch and Basques were whaling in Trygghamna, indicating that there were a lot of whales in this area (Mathisen, 1956). The remains of a Russian settlement are still visible at in the inlet of Trygghamna (fig 3c). It is considered to be dated to the beginning of the 18th century (Mathisen, 1956; Storå, 1989) and is believed to have been built between two important resource objects; Alkehornet, which is a big bird cliff with easy access to food, and the front of the glacier Kjerulfbreen which in the 18th century was located closer to the settlement than at present (Storå, 1989).
The Arctic, in particular Svalbard, has been a fascinating place to explore for centuries. Many countries have conducted research in this remote place, Sweden in particular, though few in
Trygghamna. The first mentioning of Trygghamna during Swedish expeditions is from the A.E. Nor-‐
denskiöld, Spitsbergen expedition, 1864, on the schooner Axel Thordsen when Nordenskiöld sought shelter (Liljequist, 1993). Seeking shelter in this tranquil fjord has been the main reason for entering it and in 1882, during the Geological Spitsbergen expedition, De Geer and Nathorst had to make a visit to Trygghamna. Though it wasn’t until 1901 the first Swedish research took place. It was when De Geer and Ringertz took the opportunity to perform scientific work while waiting out a storm. De Geer returned in 1910 a geological excursion to Spitsbergen, and studied the upright Silurian strata and the large glacier and performed a more detailed map of the Isfjorden area (fig 4) (Liljequist, 1993).
The geology in this fjord is rather complex. Trygghamna lies within the West Spitsbergen Fold Belt (WSFB), which consists of both sedimentary and metamorphic bedrock, caused by intensive tec-‐
tonic movements like folding and faulting (fig 3d). The fold belt is not a result of a plate collision, but is thought to be part of an intra-‐plate structure which developed 54-‐45 ma (Eocene) when sea-‐floor spreading was occurring in the Labrador Sea, Baffin Bay, North Atlantic and Eurasian Basin and Greenland acted as a separate plate and drifted towards Svalbard (Dallman et al., 2015).
Figure 3 a) Map of Isfjorden with Longyearbyen marked with a red dot. The study area Trygghamna is marked with a red rectangle. b) Satellite image from Norsk Polarinstitutt with the current distribution of the glaciers and their terminus in Trygghamna. The white outlines in the fjord are the historic glacier front margins. The outer most have been defined from the map made by De Geer (1910) and the other two from maps by Norsk Polarinstitutt (1936, 1968). c) The re-‐
mains of the Russian whaling settlement in Trygghamna. d) Värmlandsryggen with its beautiful folding and faulting.
(Photo ÅW).
c d
2.4 The glaciers of Trygghamna, their growth and retreat
According to the map made by Gerard de Geer in 1910 (fig 4), the glaciers of Trygghamna were all connected. On the map, there is only one name marked on the glacier – Kjerulf glacier (Kjerulfbreen).
On the map from Norsk Polarinstitutt (see appendix) the three larger glaciers Protektorbreen, Har-‐
rietbreen and Kjerulfbreen were connected but Kiærbreen was disconnected and the small lake Lovénvatnet is marked on the map (fig 5). On this map the names of all the glacier is written though that is likely a later addition. These glaciers and lake is not mentioned in literature until 1954-‐1955 (Orvin, 1958).
Figure 3b shows the glacier front today and how it has retreated since 1910, when de Geer created his map of the entire Isfjorden area (fig 4, 5, 6a & b).
Figure 4 Historical map of Isfjorden area made by de Geer in 1910. Trygghamna and the glaciers are highlighted in a separate frame. As can be seen from the map, the area was highly glaciated and it was considered to be only one glacier, Kjerulfbreen.
Figure 5 Map B9 from Norsk Polarinstitutt created from aerial photographs taken 1936-‐1938. On this map the nunatak Knuvelen and the lake Lovénvatnet is visible. Also the four different glaciers have their names, though that is likely a later addition to the map.
Figure 6 a) A view over Trygghamna in 1936, aerial photograph taken by Norsk Polarinstitutt. Knuvelen was a nunatak during this period and all three glaciers were connected. Kiærbreen reached all the way to the moraine.
b) Trygghamna in summer 2014, picture taken from Protektorbreen, facing NE. All glaciers have retreated and Knuvelen has a clear trimline, indicating the height of Protektorbreen in the past.
a
b
2.4.1 Protektorbreen
“The protector against the ocean off Trygghamna - the protected harbour”
(Hoel, 1942)
Protektorbreen, 78°10’-‐13°40’, lies west and north of Protektorfjellet (fig 3a). It is first marked on a map as an individual glacier in 1955 (Hoel, 1942; Orvin, 1958). The shape of Protektorbreen (fig 7a) is not the usual shape of a valley glacier, making it difficult to estimate its volume. It is <3 km long and
<2 km wide at its widest part. The more elongated part of the glacier lies in an SW-‐NE direction with a NE ice flow. The wide part of the glacier has a more direct ice flow from west to east. The northern part of the glacier reaches nearly beyond Knuvelen, and has a small accumulation area on the west-‐
ern side and the ice flow is in a NW-‐SEE direction. Otherwise the larger accumulation areas lies in the southern part of the glacier, a smaller in the east and a larger on the west side, located on a higher elevation, both on elevations above 350 m a.s.l, and is only connected by an icefall.
2.4.2 Harrietbreen
“Harriet Wedel Jarlsberg, 1846-1926. Contributor to the Norwegian Spitsbergen expeditions”
(Orvin, 1991) Harrietbreen (fig 3a and 7b), 78°15’-‐13°20’, is a tributary glacier to Kjerulfbreen, north of
Protektorbreen. It is first marked as an individual glacier in 1954 (Orvin, 1958). The shape is almost circular, with an ice flow from west to east. It is a highly crevassed glacier and the only one terminat-‐
ing in the fjord, a so-‐called tidewater glacier. Though, it is retreating and it is only a part of the glacier tongue ending in the fjord, about 2/3 is terminating on land.
a
b
Figure 7 a) Protektorbreen with the higher accumulation area in the west, connected by an icefall. b) Harrietbreen with its crevassed surface and calving front.
a
2.4.3 Kjerulfbreen
“Theodor Kjerulf, 1825-1888. Norwegian geologist, professor at the University of Oslo”
(Orvin, 1991)
Kjerulfbreen (fig 3a and 8a), 78°10’-‐13°30’, was believed to have an area of 50 km2 in the 1940’s (Hoel, 1942). This number likely included Harrietbreen and Protektorbreen (fig 7a and 7b) and not only the single valley glacier that is today. The length of the glacier is 5-‐7 km, depending on where the divide is located on the higher elevation, in some pictures it looks as Geologpasset cuts it off from the higher accumulation area. This needs to be further investigated. Though, Kjerulfbreen has two distinct accumulation areas at an elevation above 350 m a.s.l. The ice flow direction from north to south, with a slight turn to SE at the snout.
2.4.4 Kiærbreen
“Elias Cathrius Kiær, 1863-1939. Contributor to the Norwegian Spitsbergen expedition”
(Orvin, 1991) Kiærbreen (fig 3a and 8b), 78°15’-‐13°40’, is considered a tributary glacier to Esmarkbreen, which flows into the fjord Ymerbukta, east of Trygghamna (Orvin, 1958). This is the smallest of the four glaciers in Trygghamna, only 1 km long and <500 m wide. During the LIA the valley glacier was con-‐
nected to the cirque glacier on the west side of the valley, which now is completely covered with debris. By that time the glacier reached all the way to the frontal moraine ridge and the glacial lake was covered with ice.
b a
Figure 8 a) Kjerulfbreen, the biggest of the glaciers in Trygghamna and b) Kiærbreen, the smallest of them. Both display-‐
ing surface structures and melt-‐out features.
3 Methods
Various techniques were used for answering the research questions for this thesis. These include the use of remote sensing data; published sources and field mapping to track glaciological changes; and map glacial geomorphology and glaciological structures. This paragraph will outline these methods, which are all well-‐established techniques in glaciological and glacial geomorphological studies.
3.1 Data sources
3.1.1 Published data
In order to reconstruct the history of the glaciers in Trygghamna, various types of published data have been used. These include descriptions of the fjord Trygghamna from old books and articles;
photographs and maps, both historical and more recent; mapped glacier terminus positions based on old maps, one by de Geer (1910) and Norsk Polarinstitutt (1936 and 2008; see appendix). The maps have been geo-‐rectified similar to the aerial photographs.
3.1.2 Aerial photographs
To map the surface glaciological structures, three aerial photographs from 2009, retrieved from Norsk Polarinstitutt (see appendix), was used for the study area. They were provided in digital format and mosaicked by Dr. Harold Lovell. The photographs were then geo-‐rectified to Universal Transverse Mercator (UTM) Zone 33N (datum: WGS 84) using ESRI ArcMap 10.3 Georeferencing toolbar. In addi-‐
tion, several historical aerial photographs from Norsk Polarinstitutt, 1936 (see appendix), were used for identifying glaciological changes, such as ice flow directions and crevasse patterns.
3.1.3 Glacial geomorphological and structural glaciological mapping
The mapping of glacial geomorphology and features on the glacier surface was conducted digitally with ArcGIS. Features were identified and digitised as shape-‐files directly onto the geo-‐rectified aerial photographs. The identification and mapping of structural glaciological features (e.g. longitudinal foliation, arcuate fracture traces and fracture traces) on the glacier surface was performed according to Hambrey and Dowdeswell (1997), Hubbard & Glasser (2005) and Lovell et al. (2015a).
Field mapping during field camp in August 2015 was used to ground-‐truth and improve the map. Identified features in field were described, photographed and sketched within the field note-‐
book.
4 Definition and characteristics for glaciers and surge-‐type glaciers, in general and Svalbard in specific
‘Absence of evidence is not evidence of absence’
(Benn & Evans, 2010)
In this part the definition of so called normal glaciers and their structures are specified in order to understand the fascinating mechanism, dynamics and behaviour of surge-‐type glaciers.
The specific behaviour of the glaciers on Svalbard in comparison to other areas are defined, at least to
what is known today.
Tunabreen (photo ÅW)
4.1 Glaciers
The definition of a glacier is an ice mass that moves due to its own weight (Holmlund & Jansson, 2003), or its ability to transfer ice from high elevation accumulation areas to lower ablation areas where ice is lost by melting or calving (Benn & Evans, 2010). Glacier ice behaves like a thick viscous fluid, which slowly and continuously deforms under an applied stress (Cuffey & Paterson, 2010).
4.1.1 Areas on a glacier
The accumulation area (fig 9) is the part of a glacier where snow crystals are preserved during at least one melt season. By the compaction and pressure of overlying snowpack the snow transforms into firn, which is metamorphosed into ice when the air-‐ or water-‐filled passageways between the grains are sealed (Benn & Evans, 2010; Cuffey & Paterson, 2010). Thus, two processes occur in the snow-‐
pack, metamorphose when snow crystals change their shape and density, and compaction when snow is compressed (Holmlund & Jansson, 2003).
The ablation area (fig 9) is the part of the glacier where snow and ice is lost from a glacier, including processes such as melting, evaporation, sublimation and calving of icebergs (Benn & Evans, 2010). The equilibrium-‐line altitude (ELA) is the boundary between the accumulation and ablation areas at the end of the mass balance year, where the accumulation equals ablation for the year. This is a transition zone where the glacier surface changes from snow, to snow patches, to ice (Holmlund
& Jansson, 2003; Cuffey & Paterson, 2010).
Figure 9 The different areas on a glacier are illustrated by the surface of Storglaciären, northern Sweden. The accumulation area is the area on a glacier where the snow is preserved and transforms into firn and ice, i.e. gains mass. The ablation area is the area on the lower part of the glacier where the glacier experience mass loss. The equilibrium-‐line altitude (ELA) is not a distinct line but a transition area between the accumulation and ablation areas (photo ÅW).
4.1.2 Thermal state of glaciers
Glaciers do not only melt at 0°C, the increasing pressure of the ice mass decreases the melting point of the ice by 0.072°C per million Pascals (MPa). This is referred to as the pressure-‐melting point of the glacier (Benn & Evans, 2010). Glaciers are classified whether the ice is at or below the pressure-‐
melting point. The temperature of temperate glaciers is at the melting point except for a surface layer of a few meters thickness. For a glacier to be so called warm-‐based, the thickness should be approximately 100 m (Fowler et al., 2001). Temperate glaciers occur mainly in temperate maritime areas with high precipitation and summer melting (Benn & Evans, 2010). Cold glaciers mean that the whole glacier is below the melting point and are frozen to their beds. For this, the ice thickness is usually less than 60 m (Fowler et al., 2001). These types of glaciers occur only where surface-‐, en-‐
and subglacial heat sources are too small to raise the ice to the pressure-‐melting point. Polythermal glaciers have a mix of both warm and cold ice and are the most geographically widespread of the three glacier types, and exhibit a wide range of thermal structures depending on the balance of sur-‐
face and subsurface processes (Benn & Evans, 2010).
4.1.3 Glaciers on Svalbard
The Svalbard archipelago hosts up to 2000 glaciers (Worsley et al., 1986), covering 59 % (35 528 km2) of the total land area and are of different types, such as cirque and valley glaciers, ice fields and ice caps (Dallman et al., 2015). Low temperatures and accumulation rates leads to low glacial flow veloc-‐
ities and thus the glaciers on Svalbard move slower than glaciers in other regions (<10 m/y) and are, in general, not highly crevassed (Hagen & Liestøl, 1993; Hagen et al., 2003).
There are both land and marine terminating glaciers on Svalbard, but two thirds are tidewat-‐
er glaciers, i.e. terminating in water with calving fronts (Dallman et al., 2015). Valley glaciers, glaciers restricted by mountain walls, terminating in the fjords are common on Svalbard. Most important about these glaciers is the important ablation process calving (Holmlund & Jansson, 2003). Tributary glaciers are smaller glaciers that flow into and merge with larger glaciers, so called trunk glaciers.
They often contribute mass to the trunk glaciers due to their own accumulation areas (Singh et al., 2011). Western and southern Spitsbergen has a huge network of valley glaciers that originates from small ice fields (Dallman et al., 2015). Ice fields are large ice bodies that cover mountains and develop in areas with generally gentle topography and an altitude that benefits ice accumulation (Dallman et al., 2015). Ice caps are like ice fields, large continuous ice bodies, though dome-‐shaped with radial flow that covers <50 000 km2. The largest ice cap in Svalbard is Austfonna in Nordaustlandet, cover-‐
ing almost 10 000 km2 (Benn & Evans, 2010; Dallman et al., 2015). Cirque glaciers are small, round glaciers that form in a bowl-‐shaped niche in mountain ridges where snow can accumulate (Holmlund
& Jansson, 2003; Dallman et al., 2015). They were common on Svalbard, but are now in general only remnants, such as dead ice or rock glaciers.
Most of the Svalbard glaciers are considered to be polythermal (Bennett & Glasser, 2009), and most of them have temperate ice underneath a cold surface in the accumulation area. The abla-‐
tion area is below the pressure-‐melting point and thus cold based, frozen to the bed (Hambrey &
Dowdeswell, 1997; Hagen et al., 2003). Though recent studies show that Svalbard glaciers have un-‐
dergone a thermal transition from polythermal to cold-‐based (Lovell et al., 2015a). The strongly neg-‐
ative mass balance in the 20th century is causing thinning and retreat (Hagen et al., 2003), though Hagen & Liestøl (1993) concluded that glaciers with accumulation areas on higher elevations are closer to a steady state than the ones closer to the coast at lower elevations.
4.2 Surge-‐type glaciers
The most interesting about Svalbard glaciers is that a substantial amount of them are of so called surge-‐type. Less than 1% of Earth’s glaciers are considered to be of surge-‐type, but they are of great importance in understanding glacier dynamics (Jiskoot, et al., 1998; Jiskoot et al., 2000; Murray et al., 2000) and also for their contribution to global sea level rise. Surge-‐type glaciers are not evenly dis-‐
tributed around the world, but tend to be clustered in certain areas (Dowdeswell et al., 1991;
Dallman et al., 2015). This type tends to be more abundant on the Northern Hemisphere than the Southern. Svalbard is considered to be a ‘hot-‐spot’ for surge-‐type glaciers. The percentage of Sval-‐
bard surge-‐type glaciers is difficult to establish and the estimates vary substantially. Jiskoot et al.
(2000) suggests that 13% of the glaciers are of surge-‐type, Bennet & Glasser (2009) believes it to be
~35% while Hagen & Liestøl (1993) estimates as much as ~90%. The reason for this difference in es-‐
timation is likely caused by the methods used and the current climate, making it more difficult for the glaciers to build up to surge events, thus difficult to identify. Glaciers might be misclassified as non-‐
surge-‐type simply because they have never been observed to surge and display no overt signs of having done so. In order to find out the extent of surge type glaciers on Svalbard, it is necessary to find indications of, if not surge events, at least highly dynamical ice flows. As Benn & Evans (2010) proclaims: ‘Absence of evidence is not evidence of absence’. In order to find these indications it is important to test various methods, including glacial geomorphology and structural glaciology.
In various literatures the term ‘surge’ is used for any dramatic increase in glacier flow. It can be fast flowing tidewater glaciers after collapsing ice shelves or rapidly advancing glaciers caused by, for example, volcanic eruptions such as in Kamchatka in the 1980’s (Dowdeswell et al., 1991). In this thesis, the term surge-‐type glaciers only refer to glaciers with an increase in ice flow velocity caused by internally driven oscillations.
4.2.1 Surge cycle
Surges are cyclic and not directly triggered by external forcing but rather internal. Internal processes though are not independent of external forcing; a climate for making it possible for a glacier to build up into a surge is essential (Sharp, 1988; Lefauconnier & Hagen, 1991). Also, glaciers exhibit a wide range of surging behaviours, and the distinctions between surge-‐type and non-‐surge-‐type glaciers may not be absolutely clear (Sevestre & Benn, 2015).
Surge events can be explained as a two-‐phased flow regime, and indicate changes in pro-‐
cesses and conditions beneath the ice that makes the glacier shift between slow (quiescent phase, fig 10a) and fast flow (surge phase, fig 10b) (Meier & Post, 1969; Murray et al., 2003). A surge is defined as an abnormally fast flow of a glacier over a shorter period of a few months to a few years, also called the active (surge) phase. This is followed by a longer, inactive (quiescent) phase where the glacier flow slow down and the front retreats and the glacier builds up again (Sharp, 1988; Dow-‐
deswell et al., 1991; Dallman et al., 2015). Surge-‐type glaciers undergo changes in both morphology and behaviour during a surge cycle. The quiescent phase is the time between surges, the inactive phase, when the glacier is moving at a normal, slow pace. During this phase ice builds up in the res-‐
ervoir area (fig 10a) located in the upper part of the glacier, which is not to be compared to the ac-‐
cumulation area. The reservoir area usually lies within the accumulation area (fig 9), where the in-‐
crease of mass results in an increase in the surface gradient (Benn & Evans, 2010; Dallman et al., 2015). This mass gets rapidly transported down-‐glacier to the receiving area (fig 10a) during a surge, the active phase (Dowdeswell et al., 1991). In a non-‐surge-‐type glacier there is a balance between the accumulation and the transport velocity to the ablation area, thus the glacier length is more or
