Saint Petersburg State University
Hamburg University
Russian-German Master Program for Polar and Marine Sciences
POMOR
Master thesis
Reconstruction of Late Quaternary paleo-current
activity on southern Lomonosov Ridge (Arctic
Ocean) and its paleoenvironmental significance
Elena Popova
Saint Petersburg / Hamburg, 2019
Scientific supervisors
Dr. Alexey Krylov, Saint Petersburg State University
Prof. Dr. Rüdiger Stein, Alfred Wegener Institute Helmholtz Center for Polar and Marine
Research
2
Table of contents
Abstract .................................................................................................................................. 5
Аннотация (Abstract in Russian) .......................................................................................... 7
1. Introduction... ..................................................................................................................... 9
2. Study area ......................................................................................................................... 10
2.1. Geography and morphology of the Arctic Ocean ............................................. 10
2.2. Arctic Ocean circulation.................................................................................... 11
2.3. The Lomonosov Ridge morphology ................................................................. 14
3. Material and methods ....................................................................................................... 15
3.1. Material ............................................................................................................. 15
3.2. Cores routine onboard ....................................................................................... 15
3.3. Laboratory work ............................................................................................... 16
3.3.1. Total organic carbon (TOC) analysis ................................................. 16
3.3.2. Grain-size analysis.............................................................................. 16
3.4. Data processing. ................................................................................................ 17
3.5. Sortable silt mean size as a proxy: background and calculation ....................... 18
4. Results .............................................................................................................................. 19
4.1. Lithology ........................................................................................................... 19
4.2. Total organic carbon (TOC) .............................................................................. 21
4.3. Grain-size distribution ....................................................................................... 22
4.4. Grain-size fractions distribution ........................................................................ 23
4.5. Silt subfractions distribution ............................................................................. 25
4.6. Sortable silt parameters ..................................................................................... 27
4.7. Physical properties ............................................................................................ 28
5. Discussion ........................................................................................................................ 31
5.1. Lithostratigraphy and age model ....................................................................... 31
3
5.2. Sedimentation rate ............................................................................................. 33
5.3. Correlation coefficients for the fractions........................................................... 33
5.4. Moving downcore correlation and sortable silt mean record ............................ 36
5.5. Elimination of IRD effect .................................................................................. 41
5.6. Cluster analysis ................................................................................................. 45
5.7. Reconstruction of the environment and activity of currents ............................. 48
5.7.1. Dynamics of the position of the current and its influence on
sedimentation ........................................................................................................... 48
5.7.2. Changes of the sedimentological environment during the covered
period ........................................................................................................................ 51
5.7.3. Water composition and climate as main factors of circulation .......... 53
6. Conclusions and outlook .................................................................................................. 54
Acknowledgements .............................................................................................................. 56
References ............................................................................................................................ 57
Appendix .............................................................................................................................. 65
The statement on thesis’ originality ..................................................................................... 66
4
Reconstruction of Late Quaternary paleo-current activity on southern Lomonosov Ridge
(Arctic Ocean) and its paleoenvironmental significance
Elena Popova
Master Program for Polar and Marine Sciences POMOR / 050406 Ecology and
environmental management
Supervisors:
Dr. Alexey Krylov, Saint Petersburg State University
Prof. Dr. Rüdiger Stein, Alfred Wegener Institute Helmholtz Center for Polar and Marine
Research
As polar areas play an important role in the global climate and related water mass circulation
patterns, it is very useful to closely study recent and past changes in Arctic Ocean circulation
characteristics. Despite this importance, the number of studies on paleo-currents is still
scarce, mainly due to the remoteness of this area. The present study reconstructs the strength
of paleo-currents activity in a channel on the Southern Lomonosov Ridge near Siberian shelf
of the Arctic Ocean and connects these data with the paleoenvironmental history.
The material was gained during Polarstern Expedition 115/2 in September-October 2018
(Stein, 2019). Three gravity cores about 4 to 6 m length were used together with surface
sediments from box-corers in order to acquire a continuous record for the profile of the
channel system. Echo-sounding Parasound data were studied as well to complement the data
obtained from the sedimentary records.
The “sortable silt (10-63 µm) mean grain-size approach” was applied as a main proxy for
paleo-current activity. The basic principle of this approach is that ocean currents cannot
transport sand fraction (63 µm - 2 mm), while particles finer than 10 µm form aggregates,
and their primary size is not present. Thus, these fractions are excluded in this approach (cf.,
McCave et al., 1995). In order to get accurate results of the grain-size distribution of the silt
fraction, grain-size was measured using a SediGraph III Plus. Mean sortable silt values are
correlated with percentages of sortable silt and the amount of the sand fraction in order to
determine core segments with the highest current sorting and the influence of IRD input,
respectively. In addition, total organic carbon (TOC) content was measured as a parameter
for organic matter burial/preservation.
All three cores were mainly composed of fine (clayey silt and silty clay) sediments, although
numerous coarser layers were revealed. TOC contents was quite low (about 0.2% in
5
average), typically for the Arctic Ocean. Based on the preliminary age model (Stein, 2019),
changes in current activities can be reconstructed from the end of MIS 7 to present time.
Sortable silt mean size experienced significant fluctuations during the whole period
indicating events with higher current strength, and decreases in energy, proving that the
channel has been a pathway for water mass transport from the Amundsen Basin into the
Makarov Basin. The current speed proxy, together with echo-sounding and bathymetry data
as well as the age model, allowed to reconstruct changes in strength of the flow in three
levels at the bottom of the channel. Thus, both lateral and vertical migration of the position
of the current was revealed. During MIS 7 and MIS 6, the current was probably strong but
relatively high above the bottom, leading to a uniform deposition. After that, it started to
deepen and move to the south-west, narrowing down the moat and enhancing deposition at
the flank during the following stages (MIS 5 to 2?). During MIS 1, the “wide-flow” mode
established again, covering the site with sediments equally.
Periodical fluctuations of the current’s strength were discovered: slightly higher speeds were
more typical during interglacials and terminations followed by decreases during cold
(glacial) periods. This succession is considered to be related to the Atlantic Water inflows,
which in turn depended on the climatic conditions. During interglacials Atlantic Intermediate
Water was probably formed closer to the north; due to an enhanced stratification caused by
ice melting it could reach further east and may have crossed the Lomonosov Ridge. During
cold stages, on the other hand, the situation was the opposite, and the migration was
inhibited.
The attempt to determine the IRD in the record showed a few IRD peaks during the end of
MIS 4 and MIS 5. Yet, the precise division of a coarse fraction into current-sorted and icerafted was not performed.
6
Реконструкция позднечетвертичных палеотечений в южной части хребта Ломоносова
(Северный Ледовитый океан) и их влияниe на палеосреду
Попова Елена Александровна
Магистерская программа «Полярные и морские исследования» («ПОМОР») / 050406
«Экология и природопользование»
Выпускная квалификационная работа магистра
Научные руководители:
Крылов Алексей Алексеевич, доцент СПбГУ, кафедра осадочной геологии
Рюдигер Штайн (Prof. Dr. Rüdiger Stein), профессор, Институт им. Альфреда Вегенера,
Центр полярных и морских исследований в Объединении им. Гельмгольца, Германия
Поскольку полярные области имеют большое значение для климата планеты и
связанной с ним системы океанических течений, важно изучать изменения
характеристик циркуляции в Северном Ледовитом океане как в настоящее время, так
и в прошлом. Несмотря на это, изучение палеотечений здесь ведётся не так активно, в
основном, из-за удалённости региона. В данной работе проведена реконструкция
активности палеотечений в эрозионном канале в южной части хребта Ломоносова
около сибирского участка континентального шельфа Северного Ледовитого океана;
полученная информация использовалась для расшифровки эволюции условий
осадконакопления. Материал был собран во время экспедиции PS115/2 (сентябрь –
октябрь 2018) научно-исследовательского ледокола «Поларштерн» (Stein, 2019). Три
колонки донных осадков длиной 4–6 м были отобраны при помощи гравитационной
трубки, для отбора поверхностных проб в тех же локациях использовался бокс-корер.
Помимо проб осадков, был использован сейсмоакустический профиль, полученный с
помощью системы Parasound.
В качестве главного «прокси» для активности палеотечений использовался показатель
«средний размер сортируемого алеврита (10-63 мкм)». Основной принцип
заключается в том, что океанические течения, как считается, не могут переносить
песчаную фракцию (63 мкм - 2 мм) на дальние расстояния, в то время как частицы
мельче 10 мкм образуют агрегаты, и их первичный размер сложно измерить. Таким
образом, данные фракции исключаются при анализе (McCave et al., 1995). Для
получения
наиболее
точных
результатов
по
гранулометрическому
составу
алевритовой фракции, размер частиц был измерен при помощи SediGraph III Plus.
Была проведена корреляция значений среднего размера сортируемого алеврита с
7
процентным содержанием этой фракции и песка для определения участков колонок с
максимальной сортировкой осадков под воздействием течения и проверки наличия
частиц ледникового разноса соответственно. Также было измерено содержание
общего органического углерода (TOC) в качестве параметра для оценки степени
захоронения/сохранности органического вещества.
Колонки сложены в основном тонкими осадками (глинистый алеврит и алевритовая
глина), хотя слои грубозернистых осадков также были замечены. Содержание
органического углерода достаточно низкое (в среднем около 0,2%), что типично для
арктических осадков. Использование «прокси» палеотечений вместе с данными
сейсмоакустики, батиметрии и стратиграфии (предварительная возрастная модель
описана в Stein, 2019) позволило реконструировать изменения интенсивности потоков
водных масс в канале с конца морской изотопной стадии (МИС) 7 до настоящего
времени. В частности, было обнаружено горизонтальное и вертикальное перемещение
позиции течения. Во время МИС 7 и 6 течение, вероятно, было сильным, но протекало
высоко относительно дна, что приводило к равномерному осадконакоплению. Затем
произошло понижение уровня течения и его движение на юго-запад, что привело к
сужению канала там и усилению осадконакопления на склоне. Во время МИС 1 снова
установился режим широко распределённого течения, приведя к образованию
равномерного слоя осадков на дне. Были обнаружены периодические колебания
интенсивности течения: более сильные скорости оказались характерными для
межледниковий и дегляциаций, тогда как для ледниковых периодов было типично их
снижение. Такой порядок связывается с режимом притока Атлантических вод,
который в свою очередь зависит от климатических условий. Во время межледниковий
Атлантические Промежуточные воды вероятно формировались ближе к северу; кроме
того, благодаря усиленной стратификации, вызванной таянием льда, они могли
проникать дальше на восток и пересекать хребет Ломоносова по каналу. Ситуация
была противоположной в ледниковые периоды, и течение вод ослаблялось.
Попытка выявить частицы ледового разноса показала несколько пиков в МИС 4 и 5.
Тем не менее, точное разделение грубой фракции на сортированные течением
частицы и ледовый разнос не было выполнено и является предметом последующих
исследований.
8
1. Introduction
Global climate change is a complex phenomenon which involves all components of
environment as they are all interconnected. Such aspects of climate change as fluctuations
of air and water temperature, composition of atmosphere, onset of glaciers and deglaciations
have been studied actively since the methods were invented (Coope et al., 1999; Lewis et
al., 2007; Mora et al. 2007; Spielhagen et al., 2004; Thomas et al., 2018). Changes in water
circulation is another feature inevitably accompanying climate cycles because warming of
the atmosphere causes the glaciers to melt increasing inflow of cold freshwater and changing
the balance in the whole circulation system. The scale of changes depends on geography,
geomorphology, ice volume, water composition (Knutz, 2008; Jakobsson et al., 2014;
Jennings et al., 2006) Polar areas are known to be important components of the global water
circulation pattern as all centers of deep-water formation were discovered to be placed in
high latitudes (e.g., Smillie et al., 2018). The present circulation of deep water as well as
surface currents in the Arctic Ocean have been studied by many authors (Aagaard et al.,
1985; Laukert et al., 2017; Rudels, 2018; Spall, 2013; Woodgate, 2012). Changes of the
circulation patterns in the Arctic Ocean in the geological past in connection with glaciations
and deglaciations, on the other hand, have been covered scarcely so far, supposedly because
of the remoteness of the area and technical difficulties of coring and drilling. There are a few
studies that reconstruct circulation regime changes here on both a million- and thousandyears scale (Andrews, 2011; Hoffmann et al., 2019; Jakobsson et al., 2007; Lantzsch et al.,
2017). All studies related to current activity dynamics, which use the approach applied in
the presented research, cover areas south of 72° N, in the Arctic Ocean particularly – only
shelf slopes zones (McCave and Andrews, 2019); an example of such reconstruction on the
southern Lomonosov Ridge (81° N) in a deep part of the ocean is presented here.
Apart from being cold and isolated, the Arctic Ocean is distinguished by its in general low
sedimentation rate of 1 cm/ka due to low primary productivity (Stein and Macdonald, 2004),
what makes the study of the past conditions via reconstructions more complicated. Being a
distinctive oceanographic feature with high energy, intermediate and deep-water currents (or
contour currents) leave a clear trace in sediments, a combination of erosion and accumulation
processes in one place called a contourite (Smillie et al., 2018; Rebesco et al., 2014; Stow et
al., 2008) As there is a gradient of energy across the slope, sediments are inevitably
accumulating along the direction of the flow resulting in a complex of layers of increased
sedimentation rate (Bianchi and McCave, 1999). Therefore, having a better resolution, deep
9
current sediments are a reliable source of information for paleoreconstructions, especially in
the Arctic (Knutz, 2008).
The Lomonosov Ridge, being the largest geomorphological feature in the Arctiс Ocean,
divides it into the Eurasian and Amerasian basins. The size of the ridge makes it a significant
barrier for water flows. That is why any channel is considered to be a pathway for deep or
intermediate water. The aim of this work is to prove the influence of currents on the studied
site - a channel in southern Lomonosov Ridge towards the Siberian continental margin - and
to reconstruct changes in paleocurrent activity as well as to determine its possible
significance to the paleoenvironment. The main approach is using the common for currents
reconstruction proxy “sortable silt mean (SS)” (McCave et al., 1995; McCave and Hall,
2006) to determine periods with a higher or lower speed of water flow. Additionally, total
organic carbon (TOC) content was used as a parameter indirectly reflecting production and
climate change in the ocean sediments (Clark et al., 1980; Henrich et al., 1989). Physical
properties such as wet bulk density and volume-specific magnetic susceptibility were used
to create an age model by means of correlation of these characteristics with the core with
determined strata age.
2. Study area
2.1. Geography and morphology of the Arctic Ocean
The Arctic Ocean can be defined as an enclosed ocean, and its boundaries are the Eurasian
continent, Bering Strait, North America/Canada, Greenland, Fram Strait, and Svalbard, and
the shelf break from Svalbard southward to Norway (Fig. 1). More than half of the total area
of the ocean (52.7%) is comprised of the shallow continental shelves (Jakobsson et al.,
2003). The shelves form shallow, broad (up to 500 km wide) shelf seas, separated by island
groups into the Barents Sea, Kara Sea, Laptev Sea, East Siberian Sea, and the Chukchi Sea
north of Eurasia. The mean depths of shelf seas are around 50 m or less, except for the
roughly 200 m-deep Barents Sea (Mauritzen et al., 2013). The large shelf areas result in a
quite small water volume and a shallow mean depth of the entire Arctic Ocean of 1,361 m
(Stein, 2008). The Arctic Ocean abyssal plains in turn make up approximately 11.8% of the
Arctic Ocean area (Jakobsson, 2002).
10
The Lomonosov Ridge, with a mean sill depth of 1870 m, separates the Eurasian Basin from
the Amerasian Basin (Fig. 1). The Eurasian Basin is divided by the Gakkel Ridge into the
Nansen Basin (maximum depth 4000 m) and the Amundsen Basin (maximum depth 4500
m). The Amerasian Basin consists of the Makarov Basin and the large Canada Basin, both
reaching down to 4000 m deep, and separated by the Alpha-Mendeleyev Ridge. (Mauritzen
et al. 2013; Rudels, 2015)
2.2. Arctic Ocean circulation
Circulation in the Arctic Ocean is conditioned by a few factors such as restricted connection
with the World Ocean and distinctive water mass properties. It has been noted that salinity
Figure 1. Map of the geographical and bathymetrical features of the Arctic Ocean. AB: Amerasian
Basin; BIT: Bear Island Trough (Barents Sea opening); BC: Barrow Canyon; CC: Central Channel;
EB: Eurasian Basin; FJL: Franz Josef Land; GFZ: Greenland Fracture Zone; HC: Herald Canyon;
MJP: Morris Jessup Plateau, JMFZ: Jan Mayen Fracture Zone; SAT: St. Anna Trough; YP: Yermak
Plateau; VC: Victoria Channel; VorC: Voronin Canyon; VS: Vilkitskij Strait; WI: Wrangel Island
(Rudels et al., 2015). Red dot marks location of the cores used in the study.
and not temperature dominates in water mass stratification in the Arctic Ocean due to very
low temperatures here (Mauritzen et al., 2013; Spall, 2013; Woodgate, 2012), what leads to
an evident halocline and makes freshwater input significant. Such conditions led to quite a
stable stratification consisting of five layers in general. The upper one is surface or Polar
Mixed Layer underlain by halocline in which strong salinity maximum is observed,
11
sometimes containing Pacific Waters; below Arctic Atlantic Water follows with
temperatures above 0°C. Then upper Polar Deep Water is placed representing intermediate
water under which various deep waters circulate in different basins (Rudels et al. 1994;
Mauritzen et al. 2013).
There are two major gateways connecting the Arctic Ocean with the Atlantic Ocean and the
Pacific Ocean: Fram Strait and Bering Strait, respectively (Fig. 2). Fram Strait is 500 km
wide and 2600-m deep (Mauritzen et al. 2013) and is the main source of oceanic water in
the Arctic Ocean and the only pathway for deep water (Swift and Koltermann, 1988).
Surface water circulation has been studied well most probably due to the presence of the ice
(Mauritzen et al., 2013; Jones, 2001; Proshutinsky et al., 2015; Rudels, 2018; Woodgate,
2012). Circulation in the Arctic Ocean is usually described as a cycle of Atlantic Water
entering the ocean through Fram Strait and subsequent changes happening to water flow on
the way as morphology and other sources of water interfere. For the surface water circulation
wind is the major factor. Low-pressure systems above the Eurasian Basin cause a cyclonic
movement, supported by freshwater input from the large rivers Ob, Yenisei, Lena, leading
to increased ice formation (Mauritzen et al., 2013) that is enriched in the Siberian branch of
the Transpolar Drift. The opposite situation is observed over the Amerasian Basin, where
the pressure is predominantly high and circulation is anti-cyclonic. Ekman transport – to the
right from the wind direction – plays an important role here creating convergence and the
Beaufort Gyre. Surface water and ice then leave the gyre into the Transpolar Drift and move
to the south through Fram Strait (Rudels, 2018).
Intermediate and deep-water circulation expectedly has a higher connection to the
topography. North Atlantic Water flows through the Norwegian Sea where it splits into two
parts, one moving to the north of Svalbard, the other – to the south, and merge, reaching the
northern part of the Barents Sea in the St. Anna Trough. This water mass moves to the east
along the continent as a boundary current circumflexing the Siberian islands. Meeting
Lomonosov and Nansen–Gakkel Ridges a branch separates and flows back toward Fram
Strait (Rudels et al., 1994; Spall, 2013). The rest continues the flow.
As the Arctic Ocean is divided by the relatively shallow Lomonosov Ridge into the
Amerasian and Eurasian basins, and the water circulation is cyclonical and restricted, there
is a difference between the water masses of the Amerasian and Eurasian basins (Ikeda et al.,
2018). The amount of Pacific Water entering through Bering Strait is not large, nevertheless,
it creates a major difference due to its low salinity and density (Rudels, 2018; Woodgate et
12
al., 2001). At the same time, the isolation is not ultimate. There are estimates that about 5 Sv
crosses the Lomonosov Ridge into the Canadian Basin as a boundary current (Rudels et al.,
1994; Woodgate, 2012; Woodgate et al., 2001). Moreover, water from the Eurasian Basin
enters the Amerasian Basin not only along the slope in a boundary current but at
topographical irregularities of the Lomonosov Ridge (Bluhm et al., 2015; Jones, 2001).
The boundary current moving westward reaches the Alpha Ridge and enters the Amerasian
Basin where it meets and merges with the cyclonically circulating water in the Makarov
Basin. The stream splits into two parts: one following the Lomonosov Ridge and creating a
gyre in the Makarov Basin and the other crossing the Lomonosov Ridge and the Amundsen
Basin moving toward Greenland (Rudels, 2018; Mauritzen et al. 2013). The water leaves the
Arctic Ocean as East Greenland Current in the western Fram Strait, where a major part of
Atlantic Water along with some Pacific Water is returned (Rudels, 2018).
Figure 2. Schematic circulation of surface water (grey arrows) and the Atlantic Layer
plus Upper Polar Deep Water to depths of about 1700 m (black arrows). The straight
arrows represent the mouths of major rivers (Jones, 2001).
13
2.2. The Lomonosov Ridge morphology
The Lomonosov Ridge is approximately 1700 km long and 50–200 km wide. Water depth
varies significantly from about 1000 m at the ridge crest to 3900–4300 at its base. The
shallowest part of the ridge (400 m) is located near Ellesmere-Greenland shelf junction,
where 2400 m-deep saddle separates it from the shelf (Piskarev et al., 2018).
The present master thesis work concentrates on the southern part of the Lomonosov Ridge
towards the Siberian continental margin, where it divides the Amundsen and Makarov
basins. Here the flank of the ridge consists of a series of basement highs stepping down to
the basin, considered to be rotated fault blocks active during the continental rifting stage of
the formation of the Eurasian Basin. The region is characterized by a complex pattern of
ridges and basins. The largest ridge extends south of 84°40’ E. To the south, this ridge splits
into two ridges and then dies away rapidly (Cochran et al., 2006).
A narrow channel is located here between 81° to 81° 30’ N and 138° to 142° E in N-W – SE direction at approximately 1600 - 1700 m water depth (Fig. 1). The studied sediment cores
were located across this channel representing three different kinds of sedimentary
environments characteristic for a current-influenced deposit: the lower flank with high
sedimentation rate (Core 11), the bottom part where the channel is supposed to be erosive
with minimum sedimentation rate (Core 13) and a mid-point phase on the slope in between
(Core 12) (Fig. 3).
NE
SW
1720 m
11
13
12
1760 m
1 km
Figure 3. PARASOUND profile across the channel (obtained during
Expedition PS 87 in 2014; Stein, 2015), and locations of cores 11, 12
and 13 (recovered during Expedition PS 115/2 in 2018; Stein, 2019).
The coring locations were selected based on the PARASOUND
survey.
14
3. Material and methods
3.1. Material
Major part of the material used in the thesis comprises three sediment cores taken by gravity
corer (GC or “Schwerelot” – SL) during Polarstern Expedition PS115/2 in SeptemberOctober 2018 (Stein, 2019): PS 115/2_11-3, PS 115/2_12-1 and PS 115/2_13-2 (the cores
will be further referred to as 11, 12, and 13 respectively). The length of the core barrel was
10 m at stations 11 and 12 and 5 m at station 13. As the top of gravity cores is usually
considerably disturbed or destroyed, (near-)surface samples were taken by a giant box corer
(“Großkastengreifer” – GKG). All three cores were located on a north-eastern slope of a
NW-SE oriented channel on the southern Lomonosov Ridge (Fig. 3) (Table 1).
In addition to material taken in 2018, particular data from gravity Core 109 taken during the
Polarstern Expedition PS 87 in 2014 (Stein, 2015) were also used. This core was recovered
from a flat area with a normal undisturbed pelagic sedimentation to the north of the channel.
Table 1. Gear, recovery, location, water depth and the number of samples of the cores used in the
study (Stein, 2019)
Station No
Gear No
Gear
Reco
Latitude
Longitude
Depth
Samples
very,
N
E
(m)
for the
сm
PS115/2_11
PS115/2_12
PS115/2_13
study
PS115/2_11-1
GKG
48
81,0878
140,9138
1698
1 (surface)
PS115/2_11-3
SL
624
81,0879
140,9133
1696
64
PS115/2_12-2
GKG
36
81,0854
140,8925
1700
1 (surface)
PS115/2_12-1
SL
643
81,0852
140,8905
1701
43
PS115/2_13-1
GKG
39
81,0819
140,8657
1705
1 (surface)
PS115/2_13-2
SL-5
401
81,0822
140,8639
1706
46
3.2. Cores routine onboard
All the work on material acquisition and analysis is shortly presented in a flow chart (Fig.
4).
Surface samples (0-1cm) were taken from the box corer surface inside an area covered by a
plastic tube of 12 cm in diameter and stored in a 100 ml plastic beaker.
Before opening, gravity cores were first run through the Multi-Sensor Core Logger (MSCL),
and fractional porosity, p-wave velocity, wet bulk density, and volume-specific magnetic
15
susceptibility were measured. The latter two parameters were used in this study to correlate
the cores with Core 109 in order to create a preliminary age model (Stein, 2019).
After MSCL runs, the cores were cut into work and archive halves. High-quality photographs
were taken of both halves. Archive parts were visually described in order to document
lithological and structural features as well as color changes (Munsell Soil Colour Chart was
used) (Stein, 2019).
Prior to sampling the work halves, locations of samples were chosen thoroughly based on
changes in color and lithology along the cores and marked with wooden tooth sticks.
Samples were taken by a 10 ml syringe and then put into a plastic beaker. Both surface and
core samples were stored at 4°С.
3.3. Laboratory work
Laboratory analysis was performed at the Alfred-Wegener-Institute Helmholtz Centre for
Polar and Marine Research in Bremerhaven, Germany, during February-April 2019.
Before the analyses, all samples were freeze-dried at -30°С; 1 g (5 g for surface samples)
out of the whole sample was weighed and ground for the TOC analysis.
3.3.1. Total organic carbon (TOC) analysis
Sub-samples of approximately 90 mg were put in ceramic crucibles (in order to prevent
contact with carbon) and treated with a few drops of ethanol to increase the wettability, and
500 µl of 37% hydrochloric acid (HCl) to remove the inorganic carbon (carbonate). After
that, the samples were heated on a heating plate to 250°С for at least two hours. In order to
enhance combustion, iron and wolfram chips were added to every sample.
Eventually, TOC was measured on Eltra CS 2000 Carbon Sulfur Determinator calibrated
with ALPHA standard 142 (AR4020) (TC = 0.88 – 0.92%). The samples along with the
standard 142 (which was placed after each row of 11 samples for an accuracy control) were
loaded into a furnace automatically and burned with a stream of pure oxygen at about 1300°С
so that CO2 would emerge. The method is based on the reaction of pure carbon to the emitted
gas (Eltra Elemental Analyzers, 2013).
3.3.2. Grain-size analysis
Considering that the main goal of the thesis is determining current activity in the past and
the main proxy for that currently is the sortable silt mean (SS) (cf., McCave and Andrews,
2019), the grain size analysis of fine fraction was to be performed with much attention and
16
precision. That is why the SediGraph III Plus which measures mass percent of particles in
size range 0.523 - 62.5 µm was used, although, such type of grain size analysis requires
sediments to be in a condition of a highly saturated suspension (Stein, 1985). Therefore,
several steps of preparation were needed to make the samples fit this condition.
The samples were prepared by adding demineralized water and shaking for at least half an
hour in order to dissimilate aggregates that could have formed during freeze-drying. After
that, the sediments were wet-sieved through a sieve with a mesh size 63 µm in order to divide
the sand fraction from silt and clay. The sand fraction was dried at 50°С and weighed so that
mass percent of sand could be calculated.
Fine fraction was collected in 5 l plastic bowls. After sedimentation the visibly clear water
was removed, and the samples were transferred into 400 ml glass beakers. In order to reduce
the amount of water in a suspension, beakers were put into an oven at 50°С. Having reached
the possible minimum amount of water in a sample without drying, they were put into 100
ml plastic beakers. They were in turn subject to evaporation again as their smaller volume
allowed for sediments to settle down so that visible water separation would happen.
This suspension with a quite high sediment concentration was analyzed using the mentioned
above SediGraph III Plus (SediGraph III PLUS Particle Size Analyzer). It calculates the
mass percent of particles using Stokes’s law on particles settling. The principle of work is
as follows: dispersed sediment goes through a transparent plastic cell subject to X-ray beam;
the beam’s attenuation is measured as a function of time and height, what allows to calculate
the distribution of particles (Stein, 1985). Measurement of the grain size during the settling
of particles is related to transport and deposition in an environment, making use of the
SediGraph suitable for the paleo-current intensity studies (McCave and Hall, 2006).
3.4. Data processing
Excel (MS Office) was used for primary data preparation and evaluation via plotting, and
minor calculations. Basic statistical parameters and tables of correlation were calculated in
Statistica (StatSoft) software. Graphs were created in Grapher 9. The free software
environment for statistical computing and graphics R (https://www.r-project.org) allowed
performing interpolation in order to divide fine fraction into subfractions at precise size
boundaries (for example, 2 µm between very fine silt and clay) as well as cluster analysis.
Such parameters as sortable silt percentage and sortable silt mean size were also calculated
in R. Maps and other graphics were created using Inkscape.
17
Figure 4. Flow chart of data processing and analysis procedure; blocks
with gray fill indicate work onboard the research vessel (Stein, 2019).
3.5. Sortable silt mean size as a proxy: background and calculation
The attempts to quantitatively describe the capability of water flow to transport particles
depending on their size have been studied for a long time (Hjulström, 1935; Ledbetter, 1986).
Firstly, only the sand fraction had been studied, then, as techniques were becoming more
advanced, the analysis of smaller particles became possible. Particles of all size range are to
some extent sorted by the water flow as the capacity of the flow to carry material is limited.
Large particles settle down quite early in the current flow while fine are carried by the current
for long distances. Size of 63 µm has been traditionally used as a boundary between sand
and mud, and sand is considered to be too large and heavy to be transported by currents longdistance; thus, particles larger than 63 µm are excluded from the current-transportable
fraction despite that very strong currents can transport them (Lamy et al., 2015; Mao et al.,
18
2018). It has been noted that silt particles less than 10 µm in diameter behave like clay: they
are cohesive and form aggregates easily (McCave et al., 1995). In order to detect flow speed,
only primary size should be considered, that is why the finest fraction is excluded. Due to
these conditions, the size window 10-63 µm is called “sortable silt”. It needs to be mentioned
that particle size does not reflect the flow direction, only strength, therefore, the term speed
is used rather than velocity which is a vector value (McCave, 2008).
The sediments prior to grain-size analysis for this kind of proxy are in most studies treated
with at least H2O2 to remove organic carbon (McCave et al., 1995) and quite often with HCl
and NaOH to delete carbonates and biogenic opal respectively (Lantzsch et al., 2017). In this
study though this was not performed because the content of organic material proved to be
rather low (see Chapter 4.2).
Sortable silt mean size has been widely used in paleocurrent reconstructions (Hoffmann et
al., 2019; Homas et al., 2006; Jonkers at al., 2015; Lamy et al., 2015; Li et al., 2019).
Nevertheless, the precise description of its calculation was rarely found in the literature (e.g.,
McCave and Andrews, 2019). Considering it to be a major parameter in the present research,
there is a need to place here a calculation method.
Data on grain size acquired with a SediGraph is represented as a table in which cumulative
mass percentages match each measured size. First, data on all particles smaller than 10 µm
should be eliminated. Then 50% quantile is found by calculating the mean between
maximum and minimum percentage. Finally, the size corresponding to this percentage is
found by means of interpolation.
Along with the mean size in a range 10-63 µm, the mean percentage of sortable silt is
calculated (SS%). It is the sum of percentages between 10 and 63 µm, which is divided by
the total fine fraction (% <63 mm) (McСave and Andrews, 2019). Representation of data in
SediGraph allows to calculate it quite easily.
4. Results
4.1. Lithology
In general, all three studied cores are composed of silty clay colored in various shades of
brown, olive and gray with thin layers of sand at different intervals basing on visual
description (Stein, 2019). Although, the grain size was determined precisely later during the
analysis, and it proved that clayey silt is met in the cores at least as often as silty clay, so by
silty clay in the description both silty clay and clayey silt are meant. Nevertheless, the feature
19
of sharp grain size change from fine to coarse particles along the cores is delivered by the
description.
Core 11
Fine gray sediments layer of 40 cm thick is observed along with olive silty clay with several
sand layers from the bottom to 517 cm. It is followed by light and dark olive gray silty clay,
mottled at 517-472 cm. A clear boundary at 439 cm indicates 20 cm thick dark gray layer
covered by a 10 cm thick fading gray silty clay. From 429 cm alternating layers of light
grayish brown to very dark grayish brown silty clay are observed up to 323 cm, having
coarser composition and laminated character of altered silt and sand in the lower part and
bioturbation traces at 338-323 cm. Above there is an 80-90 cm layer of yellowish and light
olive brown silty clay with few thin sandy layers followed by a pattern of intervening
relatively thin brown or very dark gray silty clay layers at 258-255, 130-121.5 and 35-24 cm.
Signs of bioturbation are seen at 250-147 cm. The upper 48 cm of 624 cm of Сore 11 is
represented by very dark and dark brown partly laminated silty clay.
Core 12
Сore 12 has similar lithological pattern although with some differences. Layers of grayishbrown and olive gray silty clay are observed from the bottom of 643 cm to 557 cm, laminated
in the lower part. Several prominent sand layers are seen between 631 and 590 cm. Above
the sediments turn into massive dark gray silty clay, changed by olive gray, light olive brown
and yellowish-brown silty clay with bioturbation marks at 426 cm. At 353-331 cm mottled
grayish brown silty clay is seen, covered by 10 cm of massive olive gray silty clay and then
laminated layers of olive gray and yellowish-brown silty clay. From 280 to 207 cm layers of
thin laminated dark grayish-brown silt and sand are seen, followed by a fill of brown silty
clay. Very dark grayish-brown and brown layers of silt and sand are observed up to 191 cm,
where they are replaced with brown silty clay. At 152 cm a layer of very dark grayish brown
and brown silty clay starts with signs of bioturbation yet slightly laminated. Massive light
olive brown and brown silt dominates at 137-74 cm, covered by thin layers of sand fining
upwards to a three cm thick layer of laminated yellowish-brown silty clay and silt. Then,
similarly to Сore 11 case, a layer of bioturbated very dark grayish-brown and brown silty
clay is met up to 64 cm, after what non-laminated olive brown and brown silty clay is
observed. Six cm of laminated olive brown silty clay follow, and upper 13 cm is comprised
of dark brown and dark gray silty clay.
20
Core 13
Сore 13 is shorter (401 cm); the lithology is quite different from the other cores, yet some
similar features are present. The bottom is filled with olive brown silty clay and sandy silt.
From 393 cm a layer of olive silty clay is observed, having several layers of sand and
yellowish-brown silt; clear sandy layers are seen at 323-321 and 339-335 cm. Above, olive
brown and olive sandy silty clay is present, then it turns dark yellowish-brown at 317 cm.
At 280 cm olive gray silty clay takes place, covered by olive brown and dark grayish brown
mottled silty clay up to 264 cm. Very dark grayish-brown and olive gray silty clay is
observed at 183-163 cm with a few coarser layers at 179-176 cm. Then yellowish brown
sandy layers are seen at 163-161 cm, followed by brown and very dark grayish brown
bioturbated silty clay till 141 cm. Above there are layers of slightly mottled light olive brown
silty clay with some sand in the lower part. From 96 cm thin olive gray and very dark brown
layers intercalate, replaced by a thick layer of dark brown and light olive brown alternations
of silty clay and clayey silt. At 24-21 cm dark grayish brown sandy silty clay is seen. The
top of the core is brown and light olive brown bioturbated silty clay becoming after 11 cm
very dark grayish-brown silty clay.
4.2. Total Organic Carbon (TOC) content
Content of total organic carbon (TOC) is quite low in all three cores with a mean of about
0.2% (Fig. 5). In Core 11 it fluctuates from 0.1 to 0.67%, the numbers are high in the lower
part of the core, where two prominent peaks of about 0.5% occur, and at the very top. In the
middle part the content is relatively low and stable, though minor fluctuations are noticed.
The TOC distribution in Core 12 is quite similar. The highest content of all cores (0.93%) is
reached in a single peak at 610 cm. It looks smoother, but it is most probably due to a lower
number of samples taken. Core 13 also has a peak of 0.5% at the surface. This core is
distinctive with a lack of pronounced high in TOC content in the lower part, rather sawshaped pattern is observed here, then above 300 cm the amplitude of fluctuations falls
significantly.
21
Figure 5. Total organic carbon distribution in the cores.
4.3. Fine fraction grain-size distribution
The compilation of grain-size distribution of the fine fraction (< 63 μm) in all samples is
shown in Fig. 6. The cores demonstrate similar shape yet vary, mostly in amplitude. There
is a small peak of 2% at about 5 μm and a higher one of 4-6% at 30-63 μm. In Core 11, fine
particles comprise relatively high percentage while the coarse peak is narrower and lower
than in the other cores. Just like in Core 11, Core 12 has a peak at 50 μm, though it is higher
(up to 5.6%). The peak of fine silt is the least prominent in Core 13, but the amplitude is the
largest. The coarse fraction peak is wide and includes finer fractions, unlike in other cores.
Such distribution indicates that the cores consist of layers with various grain-size, and the
presence of very coarse silt is significant. Core 13 shows the highest diversity in sediments
of different size.
22
Figure 6. Mass percent grain-size distribution (fraction <63 μm) in the
cores.
4.4. Grain-size fractions distribution
Detailed analysis of grain size allowed to study the distribution of particles of different sizes
in the cores. Generalized Udden-Wentworth classification (Folk, 1957) was used (Table 1).
As the research is concentrated on the study of paleo-current activity, the emphasis was made
on the silt fraction as these are current-transportable particles (McCave et al., 1995).
Table 1. Grain size classification used in the study (Folk, 1957)
Fraction Sand
Size, μm
>63
Very
coarse
silt
63-31
Coarse
silt
Medium
silt
Fine silt
Very fine
silt
Clay
31-16
16-8
8-4
4-2
<2
A major part of all three cores is composed of clayey silt and silty clay (Fig. 7). Although,
such types of sediments as silty sand and sandy silt are met as well in smaller quantities
varying in every core.
23
Figure 7. Sand, silt, and clay distribution in the cores.
Core 11
Core 11 is mainly composed of clayey silt with the silt fraction ranging from 31 to 73%; in
12 samples the amount of clay reaches up to 64%, so that silty clay is met. Silty sand is
present in the core at depths 167-166, 345-344, and 605-604 cm, where sand comprises about
half the percentage. Changes along the core are not dramatical, except noticeable sand
content peaks which are accompanied by significant fall in clay. Yet in one case (167-166
cm) the clay fraction percentage stays at 26%, what can indicate bad sorting in the layer.
Another example of a change is observed in clay and silt content: in the middle part of the
core (440-120 cm) the period of fluctuation is much higher than above or below; the
amplitude though is stable.
Core 12
The silt fraction again prevails in Core 12, fluctuating from 19.4 to 69.7%, so the sediments
are clayey silt. Similarly to Core 11, silty clay is observed when clay content exceeds half
the percentage and reaches up to 61.3%, what happens in 11 cases. In layers at 88-87, 189.524
188.5 and 580-579 cm silty sand is present with a maximum sand content of 70.7% in the
lower part, the amount of clay in this type of sediments is still quite high though. There is
one sample at 570-569 cm where silt prevails, yet sand and not clay has a second great
percentage, comprising sandy silt which is met only in this core.
The change of the sand fraction along the core is slightly different from that of Core 11: the
mean amount of sand is larger (11% against 8% in case of Core 11), and peaks are not that
sharp and are surrounded by smaller peaks. A saw-like pattern of silt and clay is seen here
in the upper meter of the core. In general, the fluctuation of all grain-size fractions is in
agreement.
Core 13
The dominating type of the sediment in Core 13 is again clayey silt with highly variable
from 29.6 to 84.4% silt content. Silt reaches up to 65.5% in silty clay which is observed only
in five samples. Silty sand is present in the core close to the top, at 2.5 cm, although the sand
fraction here does not exceed silt greatly (44.6 against 36.3%). The fluctuation of grain size
composition in this core is quite different. There are no clear highs in sand content, its
distribution is more uniform, except for a single peak on the surface and two bulges at 320200 cm. Changes in composition make high period fluctuations below 280 cm and chaotic
low amplitude fluctuations above. Moreover, sand, silt, and clay are apparently less
connected to each other, unlike in other cores.
4.5. Silt subfractions distribution
As it was mentioned before, the analysis of fine fraction composition is necessary for the
study of current influence on the site. This is why the detailed information on the distribution
of all subfractions of silt (very coarse, coarse, medium, fine, and very fine) is available.
Considering mean content and maximum and minimum values of all subfractions (Fig. 8),
the character of distribution is quite similar in cores 11 and 12: the trend of mean percentage
is fining with fine silt is peaking at 14% in Core 11 and 13% in Core 12, followed by very
fine silt which has only 0.4% less in both cases. Coarse silt has the smallest mean value
(about 5%) but the largest amplitude fluctuating from 0 to 28.6% and 26.7% in cores 11 and
12 respectively, what indicates its low content in general along with sharp coarse fraction
peaks seen earlier in the sand record. These data support the description of cores 11 and 12
as fine in general with several coarser layers.
25
Core
13
has
quite
different
characteristics. The mean content is
approximately the same (10-11%)
for every subfraction, which can
point to the bad sorting in the core.
Very coarse silt again has the largest
variation, yet unlike in other cores,
the maximum is leaning towards
coarse particles reaching almost
47% in the coarsest part. This is the
evidence for this core to contain the
greatest amount of coarse material
and to be in general less sorted
comparing to the others.
Fig. 9 shows changes in the silt
content
along
all
cores;
the
percentage is recalculated so that
100% is a silt content in total. In all
cores changes in very coarse silt are
the most prominent, in Core 13 to a
less extent though.
The fluctuations repeat the pattern of
grain size fractions distribution:
periods with high amplitude agree
with such in Fig. 7. Apart from
numerous coarse layers, coinciding
with sand peaks, small short finings
of sediments are clearly seen, for
example, in core 11 at 427.5-426.5
cm and 128-127 cm, in core 12 at
326-325 cm. There is one case of an
increasing amount of very fine silt up Figure 8. Box-whisker plots of silt subfractions in each
to 44% without major changes in the
core.
very coarse silt fraction at 268-267 cm in Core 11.
26
Such distribution demonstrates the dominating role of the fine fraction in the cores (fine silt
and very fine silt) with high coarse peaks occurring rapidly, and only a few periods of finer
material.
Figure 9. Distribution of subfractions of silt according to the
classification by (Folk, 1957).
4.6. Sortable silt parameters
Being a major proxy for the strength of water circulation in the past, sortable silt mean
(SS, SS mean) as well as the percent of sortable silt (SS%) were subject to a detailed study.
According to (McCave et al., 1995; McCave and Hall, 2006), a high correlation between
these two parameters can indicate that sediments were current-sorted to such an extent that
they hold a record of current speed.
The total content of SS and SS% (Fig. 10) varies in all cores. Comparison of mean SS gives
a slight trend increasing towards Core 13: in Core 11 it is only 20.8 μm, in Core 12 – 22.3
μm, while in Core 13 – 24.3 μm, which again points to a coarser content of the latter core.
Mean value is much closer to the minimum of about 14 μm in all cores than to the maxima
of up to 43.5 μm in case of Core 12, indicating occasions of rare highs rather than a uniform
27
Figure 10. Box-whisker plots of 𝑆𝑆 and SS%.
distribution. Percentage of sortable silt fraction follows the same pattern, reaching 22.4% in
Core 11 and 24.5% in Core 12 but 32.4% in Core 13.
Variation of the parameters along the cores (Fig. 11) is quite high in all cores. In Core 11,
the sharp maximum is reached close to the bottom by both variables; above they fluctuate
without a definite period. In general, the shape of the curves is similar, yet there are
differences, for instance, at 280-260 and 200-150 cm. The correlation coefficient between
the parameters is 0.83, which is quite high.
Despite the visually different shape of Core 12 curve, some patterns noted in core 11 are met
here as well, including a section at 150-100 cm. The curves in this core are smoother,
especially at depth 560-330 cm, what can be related to the flattest part of 11 core’s curve
(580-460 cm). The very bottom has, similarly to the Core 11 case, short high amplitude
fluctuations. Above 300 cm variations seem more or less periodic with constant amplitude
and a slightly decreasing trend towards the surface. The correlation coefficient is identical
to that of core 11.
The curves of Core 13 seem outstanding first with their high amplitude and second with a
lower connection between parameters (correlation coefficient is 0.78). Indeed, lower 100 cm
are presented by high-periodic fluctuations. Above there is a low amplitude curve section up
to 130 cm, distinctive with low correlation between SS and SS%; it corresponds to similar
areas in other cores. Periodic fluctuations in the upper part are again stronger.
Basing on the information above, it can be concluded that SS fluctuates actively in each core
from about 14 μm to slightly over 40 μm. Similar patterns allow correlating all three cores
with each other, at least in the lower parts. Calculation of correlation coefficients indicates
28
a high role of currents in cores 11 and 12, and rather lower in Core 13. Although, as the
coefficient was calculated for the whole record, and the curves reveal dynamics in the
connection between parameters along the cores, there is a need in distinguishing the
importance of current sorting in different parts of the cores.
Figure 11. Calculated parameters percent of sortable silt (SS%) and sortable silt mean
(SS mean).
4.7. Physical properties
In order to receive preliminary data about the cores and later create an age model, physical
properties such as fractional porosity, p-wave velocity, wet bulk density (density) and
volume-specific magnetic susceptibility (MS) were measured (F. Niessen/Stein, 2015,
2019). The latter two parameters proved to be sufficient for correlating of the cores with
Core 109.
29
Both parameters fluctuate actively, forming patterns which are observed in each core (Fig.
12). Starting from the bottom, the first pattern is high amplitude fluctuations of density along
with rising MS, for example, at 631-550 cm in Core 12 or 630-590 cm in Core 109. Above
there is a peak in MS and a notable discrepancy between the parameters (570-500 cm in core
11, 290-130 cm in core 13). Then another peak in MS is seen followed again by divergence
with prominent MS low (330-210 cm in Core 12, 120-60 cm in Core 13). Two higher peaks
of both parameters and one low in between are observed in cores 109 (260-200 cm), 11 (200160 cm), and 12 (110-80 cm). Then increasing trend of both parameters is seen (160-110 cm
in Core 11), and then they go slightly down and stay stable. At the top, there is a decreasing
trend for both MS and density after a small crest. Fluctuations in the upper part of Core 13
do not express any particular pattern, except for probably a peak at 1-0 cm.
Figure 12. Wet bulk density and volume specific magnetic susceptibility for cores
11-13 (Stein, 2019) and 109 (Stein, 2015).
30
5. Discussion
5.1. Lithostratigraphy and age model
Paleoceanography studies reconstruct events that happened at some period in the past,
therefore, proper age control is required here. As an absolute age determination was not
performed for the cores in this study, the method applied here for dating is the correlation of
the cores with the core with known ages of strata basing on their lithology and physical
characteristics.
Results of the lithological description of the cores agree well with the general structure of
sediments for the Arctic Ocean (Clark et al., 1980) describing them as a sequence of
dark/brownish and grayish/olive layers. The brown layers are interpreted as sediments
deposited during relatively warm periods (interglacial/interstadial), gray indicate periods
with lower temperatures (glacial/stadial).
Along with lithological features, changes in physical characteristics were used to enhance
precision in the determination of boundaries between the stages. As it was emphasized
above, magnetic susceptibility and wet bulk density curves create patterns observed in all
three cores.
The result of the correlation is shown in Fig. 23 (Stein, 2019). The data on Core 109 is taken
from (Stein, 2015; Stein et al., 2017). First, Core 11 was related to Core 109 using physical
properties curves, then patterns of both lithological features and physical properties
characteristics in cores 12 and 13 were matched with those of Core 11 in order to determine
the MIS boundaries in all cores.
The first unit from the bottom is characterized with presence of numerous sandy layers; there
is a peak in MS which was used as a boundary between MIS 7 and 6 (190 ka) (all ages of
MIS boundaries according to Lisiecki and Raymo, 2005). The part above contains prominent
gray layers in cores 11 and 12 and olive layers in Core 13; it is bounded by a large minimum
in MS considered to be the base of interglacial MIS 5 (130 ka). This part is very thick in all
cores and marked with several brown layers including especially a big one in the upper part
of cores 11 and 12. Two peaks of density and MS comprise MIS 4 (74-59 ka), what agrees
with lithology as brown layers are absent here. The next unit is MIS 3 with thick brown
layers; the upper boundary is a minimum in both physical properties (28 ka). Then there is
plateau accompanied by a fine sediments unit with thin coarse layers, which is considered to
31
be MIS 2. It is followed by a short upward trend in physical properties indicating the start of
MIS 1 (13.5 ka). The color is also darker here.
MIS 4, 3, and 2 were not present in Core 13. The assumption is that they were eroded by
strong water flows or barely deposited for the same reason. Nevertheless, the sharp erosive
contact was not noticed, probably due to bioturbation, so the boundary between MIS 5 and
MIS 1 is less precise here than in other cores.
Figure 13. A preliminary age model with isotope stages MIS 7 to MIS 1, based on the
correlation of cores 11, 12 and 13 to Core 109 and to each other using lithology and
physical properties, is presented. Coarse-grained layers are indicated as dotted intervals.
Intervals marked by light orange shading highlight lower MIS 5 including Termination II
(Lototskaya and Ganssen, 1999; Moseley et al., 2015). Black numbers at the right are
ages of MIS boundaries in ka = thousands of years BP (cf., Lisiecki and Raymo, 2005).
The numbers surrounded by green rectangles are thicknesses (in centimetres) of the MIS
5 and MIS 6 intervals (Stein, 2019).
32
5.2. Sedimentation rate
Based on the boundaries of marine isotope stages, the linear sedimentation rates were
calculated. The results are represented in Table 2 (cf., Stein, 2019). As the boundary between
MIS 5 and MIS 1 is not determined precisely, the numbers for the sedimentation rates are
rough.
Table 2. Sedimentation rates for the cores Figure 14. Age plotted against depth.
during MIS 6 to MIS 1 (in cm/ka-1) (cf.,
Stein, 2019)
MIS/core
109
11
12
13
1
4.3
3.8
1.4
0.8
2
4.7
2.7
2.1
0
3
2.3
2.2
1
0
4
4.2
3.1
2.1
0
5
4.1
4.4
3.6
0.7
6
1.7
2
4
3.3
Table 2 and Fig. 14 reveal the general picture of sedimentation in a contourite system
(Faugères and Mulder, 2011; Rebesco et al., 2014): higher sedimentation rates and active
deposition are observed further away from the center of the channel and the current’s main
trajectory (Core 109 at the top of the drift and 11 on the lower flank of the channel), while
in Core 13, taken from a bottom part of the channel, the sedimentation is limited, and a
period of non-sedimentation or erosion is observed during MIS 4-2. As for changes in time,
the sedimentation rates are relatively high during MIS 6-5 (especially in cores 12 and 13),
then at about 70 ka (beginning of MIS 4) the curves become flatter meaning that
sedimentation visibly reduced. That indicates more or less uniform deposition in all three
cores in the first part of the studied period and change in the sedimentation mode later when
the central area of the channel underwent scarce sedimentation and upper flank, on the
contrary, gained more sediments.
5.3. Correlation coefficients for the fractions
Grain-size analysis proved that there is a general coarsening gradient from Core 11 to Core
13, that is from the outer part of the channel slope towards the central part. Apart from
observing a total change in the coarse or fine fraction, there is a need in tracking the
differences of the relation of fractions to each other, so that the most important components
could be revealed. Moreover, high correlation between the fractions can indicate similar
33
nature of the transport process (Revel et al., 1996). In order to perform that, correlation
̅̅̅, and standard deviation, used as a measure of sorting (Revel
coefficients for all fractions, SS
et al., 1996), were calculated. Major meaningful (p < 0.05) positive connections are shown
in Fig. 15.
In Core 11, very fine silt is tightly connected to fine silt, which in turn correlates with
medium silt and sorting, which proves the dominating role of the connectionfine fraction in
the core. Yet, clay is less significant here. Another cluster seems to have a totally opposite
character: it is formed by coarser elements such as very coarse silt and sand, which slightly
correlate with SS. This is not expected as very coarse silt comprises particle with sizes 3163 μm while mean SS for this core is only 20.8 μm, and the sand fraction supposedly consists
of ice-rafted detritus to some extent. At the same time, the sand fraction has an insignificant
negative correlation with the medium to very fine silt range.
The picture for Core 12 is alike, although there are several distinctive details. Clay is more
connected here to very fine silt; the sand fraction correlates to SS even stronger than in Core
11 (correlation coefficient is 0.72 against 0.66), while the connection to very coarse silt is
lost. Negative connection is met even for a larger number of silt subfractions, from coarse to
very fine.
The connections in Core 13 are in general weaker. Only the compounds SS–very coarse silt
and fine silt–very fine silt remain stable. Here, sand does not correlate strongly positively
with any fraction. The weak negative correlations are in turn observed with all silt-size
particles.
One of the features in this sequence is the detachment of the sand fraction towards Core 13:
it has less and less connection with other fractions. Accordingly, a number of negative bonds
grows. This can indicate that the sorting is very bad, and presence of sand becomes more
“unnatural”, so apparently either the source of this fraction is discrepant, or the composition
of the sediments has been altered as a result of some process.
According to (Hass, 2002), an assumption that the sand fraction is represented only by icerafted material and the main factor influencing the silt fraction accumulation is currentsorting requires minimal correlation between these two fractions. In the case of the current
study, a correlation of the sand fraction with silt is negative and relatively large in Core 11,
slightly stronger in Core 12, and the smallest in Core 13. Such a situation creates an
implication that the coarse fraction is not necessarily ice-rafted and/or the silt fraction is not
34
Figure 15. Connections between the fractions, sorting (standard
̅̅̅ basing on correlation coefficients (solid line shows
deviation) and 𝑆𝑆
coefficients values 0.75-1, dashed – 0.5-0.75, the borders are based on a
modified Chaddock’s scale (Chaddock, 1925)).
current-sorted. Assuming that IRD input was equal in all cores as they are located quite close
to each other, the changes in the silt fraction acquire more weight in the correlation. In Core
13 the correlation between sand and silt fractions is the smallest; thus, the silt fraction here
was the most prone to current sorting. Stronger correlation in Core 12 points at a more active
sorting here than in Core 11, which was placed higher on the channel flank. At the same
time, presence of some correlation between sand and silt does not allow to consider sand to
be composed of IRD only, though it can contain such particles (Hoffmann et al., 2019).
Moreover, the periods of IRD deposition could just coincide with enhanced currents modes.
Considering this, the attempts to eliminate the IRD effect should be taken with extreme care.
35
5.4. Moving downcore correlation and sortable silt mean record
It was proven that under a current’s action, the sediments become well-sorted. Several
attempts have been made to find a parameter that can estimate a measure of current sorting
of the sediments. Some authors claim that positive correlation between SS and SS% is an
indicator of sorting by currents (McCave et al., 2017). Use of the correlation between SS
and SS% for this purpose gives satisfying results being applied to various data sets (Andrews
et al., 2018; McCave and Andrews, 2019). Calculation of this parameter for the whole core
can give an estimation of how much it was affected by currents, and gives an opportunity to
compare it with other cores in different locations.
Although, the environment and factors of sedimentation mode change with time even in one
point, what created a need in the estimation of sorting by currents in different parts or along
a core. One of the possible ways to do that is to calculate a correlation coefficient separately
for each MIS. Considering different sedimentation rates during every period, the number of
samples varies significantly, making such an analysis not very reliable. The solution to this
might be the calculation of running (moving) downcore correlation coefficient, which
creates a continuous record of correlation between SS and SS% (McCave and Andrews,
2019). It allows to compare sorting in segments of a core and discover when the currents
reworked the sediments most and when their influence was not that great. The window of
correlating points was chosen basing on the length of the cores, sampling resolution and
shape of a graph, which should be neither too smooth nor have too high amplitude of
fluctuation.
At the same time, considering SS alone to be a parameter sufficient to indicate currents
influence, the correlation between SS and SS% can be used as a factor determining whether
it is appropriate to use SS as a current speed indicator or not. As a result of calculations,
coefficient of 0.5 was suggested to be used as a borderline, making SS a proxy of currents
action not applicable in the segments of the cores where the correlation coefficient is less
than 0.5.
Fig. 16 shows the running correlation coefficient in cores 11, 12, and 13. Several runs with
various windows proved 7 points to be the optimal width. The results vary in all cores. One
whole part from the end of MIS 5 to the beginning of MIS 2 showed a low correlation in
Core 11. Glacial periods were to the most extent subject to a low correlation in Core 12,
although the parameter fluctuates actively during MIS 6 and 5, and is very close to 0.5 during
36
MIS 3. During both MIS 6 and 5 in Core 13, the correlation tends to zero and below, yet in
between there is a part with very high correlation.
Basing on that, MIS 7 proved to be the most stable in means of a high correlation between
SS and SS%, indicating the highest degree of sorting by currents. MIS 6 shows a decreasing
trend from Core 11 to Core 13: in the first one the correlation is not too high yet above 0.5,
in Core 12 it drops to zero twice, in Core 13 the coefficient is below 0.5 at the beginning of
MIS, but at the end it rises high. The shape of the graph during MIS 5 looks quite similar in
the cores: it is high at the beginning of the period and then gradually decreases. In Core 13
there is a high at the end of this MIS, but it is necessary to consider that the boundary between
MIS 5 and 1 here is not very accurate. Low correlation is observed in MIS 4 in both cores
where this period is present; the same is applicable for MIS 3, though the coefficient is higher
in Core 12. There is a low in Core 12 during MIS 2, but in Core 11 the correlation is, on the
contrary, peaking. The correlation coefficient is above 0.5 in all cores during MIS 1,
although unlike Core 11, where it decreases in time, in cores 12 and 13 it increases.
Such a comparison makes an implication of different extent of current sorting varying with
time in each core, what is seen clear, especially in MIS 6. There is a need to notice that, as
it was mentioned before, sampling resolution varied in all cores and along a core; therefore,
a number of samples per MIS varied significantly making a comparison rather rough,
particularly in parts MIS 4-1 with a quite low sedimentation rate. Nevertheless, such an
analysis can be and was applied to accompany the record of SS in order to evaluate the
reliability of this parameter. Segments of the cores with a low correlation between SS and
SS% were not excluded from the further analysis of SS, yet the evidence of such a low
possible current sorting was considered. Fig. 17 shows SS records which include running
downcore correlation data to emphasize the periods when the sediments were less affected
by the currents.
37
Figure 16. Running downcore correlation between 𝑆𝑆 and SS%, the areas with gray
shading indicate correlation coefficient > 0.5.
38
12
11
13
Figure 17. Running downcore correlation with the 𝑆𝑆 record.
39
Sortable silt mean record (Fig. 18) in Core 11 shows high peaks corresponding mostly to
transitional phases or interglacials: high numbers (above 25 μm) are reached during MIS 7,
at the beginning and middle of MIS 5 and the beginning of MIS 1 as well as at the transition
of MIS 4 to MIS 3, though downcore correlation does not indicate SS during this period as
current-sorted. All the peaks in Core 12 repeat the picture for Core 11. Resolution of MIS 7
is higher here so that high amplitude fluctuations are seen. Core 13 is notable with higher
amplitude and period of fluctuation, especially in MIS 7 and MIS 5, which indicates very
strong flow during interglacials and rapid changes in speed. The preliminary age model for
this core indicated extremely low sedimentation rate during MIS 4-2, the reason for this is
most probably a too high speed of flow for the particles to deposit or even erosion during
the period of a strong current, although no sharp contact was found. Bad current-sorting is
suspected in a great part of MIS 6. The middle part of MIS 5 is supposedly also not currentaffected as the correlation between SS and SS% is quite low. This period is marked with
several climate changes and events, so changes in speed of currents could correlate with
them.
Figure 18. Distribution of 𝑆𝑆 in the cores; glacial periods are shaded gray.
40
Comparison of the numbers for SS with the published data from different areas proves the
sorting in this study to have both significant strength and high amplitude. SS in this study
fluctuates from 14 to 40 μm while for example in the North Atlantic it changes only from 16
to 25 μm (Jonkers et al., 2015), in the South China Sea from 16 to 23 μm (Li et al., 2017)
and 14 to 21 μm in the Labrador Sea (Hoffmann et al., 2019). Only SS record from the Drake
Passage reaches close: 16-37 μm (Lamy et al., 2015), what, considering the power of water
going through the Passage, evidences reasonably strong currents in a channel on southern
Lomonosov Ridge.
5.5. Elimination of IRD effect
As the study area is located close to the North Pole, it is surrounded by places of iceberg
formation as well as seasonally and perennially covered with the sea ice (Obrochta et al.,
2014; Stroeve and Notz, 2018). Both these factors of ice presence in the ocean influence
sedimentation character as floating ice can be a source and transport agent of sediments.
Being a solid substance, the ice floes, unlike sediment transport within the water mass, may
entrain and transport large and heavy particles, and deposit them after melting. The
terrigenous particles which were transported and deposited to the bottom of the ocean are
called “ice-rafted debris (detritus)” or IRD (e.g., Bischof et al., 1996; Darby et al., 1997).
Therefore, a presence of coarse-grain sediments in the core from high latitudes quite often
indicates a presence of ice there. At the same time, the activity of strong currents can cause
winnowing and lead to a formation of coarser layers, not to mention that sometimes currents
can carry particles larger than clay and silt (Lamy et al., 2015). At the same time, silt can be
carried by the ice too, and in great amounts. This poses a problem of distinction of one source
of coarsening from another while trying to estimate a degree of current sorting in an Arctic
core.
There have been several attempts to distinguish and/or eliminate IRD in sediments, for
example, using end-member modeling (e.g., Jonkers et al., 2015; Hoffmann et al., 2019;
Wang et al., 2018). Some authors though claim this analysis to be ineffective for the ice
effect elimination in current-sorted sediments (McCave and Andrews, 2019). Another
approach is taking the whole sand fraction as ice-rafted due to the implication that such large
and heavy particles cannot be transported by currents (Hoffmann et al., 2019) and then using
deviations from a regression function created by plotting the sand fraction content against
SS (Hass, 2002). Fig. 19 shows such graphs and equations for each core. The function
describes the correlation between the sand fraction and SS. Samples which lie on the
41
regression line were deposited under “average” conditions when IRD variations were not
high, while scattering around the line indicates moved equilibrium with either high IRD
input or strong current sorting.
Inserting sand content into the equation as x allows to calculate potential SS (SSpot)
supposedly showing the SS record without current influence. Subtraction of SSpot from SS
gives the so-called delta SS – the equivalent of SS, only without IRD influence. There is a
need to notice that during the current study no measurement of the upper limit of sand was
made. Based on visual observation, 2 mm were used as the borderline.
Figure 19. 𝑆𝑆 in phi scale plotted against sand percentage,
and a trend line.
The results of the calculations for every core are shown in Fig. 20. In all cores, the curve of
delta SS has absolutely the same shape as SS, only with slightly higher amplitude and a few
(4-5) μm lower. SSpot fluctuates only within 0.79-1.33 μm. In order to compare delta SS and
SSpot, the ratios of their variances were calculated. In Core 11 this parameter equals 8.5, in
Core 12 it is 7.9 while in Core 13 – 10.4, what indicates that major part of the silt fraction
fluctuation is current-induced and to a higher extent in Core 13. Surprisingly, the ratio in
Core 12 is somewhat lower than in Core 11. Considering that the variance of delta SS is
gradually falling towards the bottom of the channel (from 0.37 through 0.36 to 0.3), the
reason is a higher SSpot variance in Core 12 (0.05 against 0.04 in Core 11). Therefore, such
42
calculations do not agree with the assumption that IRD input was the same in all cores which
is unrealistic. Although, high variation of the coarse fraction does not necessarily mean
increased or reduced IRD input in one particular core.
Several small peaks in SSpot (it is given in phi-scale, so smaller numbers mean greater size)
were noticed in all cores’ records. They are in general coincide with sand peaks. A fall and
then a sharp rise was observed during MIS 4 in cores 11 and 12. This peak could be a true
IRD-rich layer, as the deglaciation phase is accompanied by active icebergs movement in
the ocean, which then melt as the temperature rises; according to the studies (e.g., Svendsen
et al., 2004), Eurasia experienced the glaciation covering probably even Severnaya Zemlya
archipelago at the end of MIS 4, so that the glacier was the closest to the studied area. The
peaks almost always follow more prominent lows. Some of the lows are though individual,
for example, at the end of MIS 6 in all cores, coinciding with low delta SS. Cores 12 and 13
have smaller falls at the end of MIS 7, which probably might be found in Core 11 has it
longer recovery of this period. Core 11 has another fall and rise in the middle of MIS 5, most
probably reflecting another IRD input during the end of one of the glacial sub-stages, such
as 5b (the precise age is hard to determine using current age model) (Spielhagen et al., 2004).
Considering that SSpot presumably shows sortable silt fluctuation as it would be without
current-sorting, very flat and smooth distribution of this parameter in all cores demonstrated
the dominant role of currents as a sedimentation factor in the area, while delta SS repeating
SS pattern only supports this. Unfortunately, the precise size range of the sand fraction was
not determined; this fact could influence on the calculations, leading to artifacts. Also,
considering that the sand fraction can be current-sorted as well as transported by the ice
while here all particles larger than 63 μm are considered to be IRD, the results of
distinguishing between IRD and current-sorted material require closer study using probably
some other techniques. The main feature that distinguishes current-influenced material from
all other types is still good sorting, that is why it should be used as a current activity indicator.
43
Figure 20. Sand percentage, 𝑆𝑆, potential 𝑆𝑆 and delta 𝑆𝑆 in the cores.
5.6. Cluster analysis
K-means cluster analysis was performed using the following parameters: TOC, SS%, SS,
sorting (standard deviation), SSpot, delta SS, wet bulk density, magnetic susceptibility, the
content of sand, very coarse silt, coarse silt, medium silt, fine silt, very fine silt, and clay. An
optimal number of clusters was six (Fig. 21), despite some of them intersect.
Figure 21. Cluster plot of all samples.
45
Each cluster had its combination of characteristics; the most distinctive are represented in
Table 3. The evaluation was based on distribution of the parameters: mean, minimum and
maximum values; grain-size column summarizes all grain-size fractions data.
Table 3. Characteristics of clusters (color scale does not match that of Fig. 21)
Cluster
1
2
3
4
5
6
TOC
middle
high
middle
middle
low
high
Grain-size
clayey sand
coarse silt
fine silt
silty clay
silty sand
fine silt
SS mean Sorting
middle
middle
very low
very low
very high
low
Density
very good middle
middle
high
very bad
low
very good very low
very good very high
bad
very low
Plotting clusters in the cores with the developed earlier age model (Fig. 22) proved the
quality of the age model as in general each MIS had its distinctive sequence of layers; the
glacial and interglacial periods also have different features. The goal of this analysis was to
classify all samples basing on the known information at the same time grouping them, and
particularly to determine the period when the accumulation in Core 13 was scarce, or erosion
happened. Every sample was considered here so that a high-resolution record was created.
MIS 7 is the only period when sediments of cluster 5 are present in all cores. They are the
coarsest, the densest, current-sorted and poor in TOC, indicating the strongest water flow
impulses. These sediments are combined with clusters 1, 2, 3 and 4, in Core 12 presence of
cluster 1 is more prominent. This is some kind of a transitional type considering its mixed
grain-size yet good sorting. In Core 13 sandy layers intervene with dense rich in organics
coarse silt of cluster 2, indicating lower energy of the environment.
MIS 6 looks similar in cores 11 and 12: cluster 6 of low density rich in TOC bad-sorted fine
silt with an indication of low current influence dominates. There are more of cluster 1
sediments in Core 13 in this period so that no sharp boundary is seen between MIS 7 and 6.
It could be due to mixing during some small-scale mass-wasting processes or bioturbation.
Similarly to MIS 7, layers of cluster 4 are observed here, only in larger amount. These
sediments are very fine in grain-size and have low SS meaning that no strong current action
happened there at that time. In all three cores, this period ends with sediments belonging to
the cluster 3, which represents very bad-sorted fine silt with no evidence of increased
current’s speed. The composition agrees with the glacial period’s environment during which
46
the cluster is seen: IRD input explains bad sorting, and weaker water stream reduces current
sorting.
Figure 22. The cores divided into 6 clusters.
MIS 5 is represented by various clusters combination depending on sampling resolution and
sedimentation rate in the core. This situation reflects numerous climate changes during this
period (Oppo at al., 2001). It starts with a layer of cluster 2 and transforms to cluster 1,
indicating a period of intermediate current influence. Then in cores 11 and 12 clusters 4 and
6 dominate evidencing a slow-down till the beginning of the MIS 4. The upper part of MIS
5 in Core 13 is different: it is built with sediments of mixed type, spreading to the surface.
MIS 4 is the most easily distinguished glacial period. It contains sediments of cluster 3,
typical for the cold period, and 1, which is characteristic for transitional periods. The sorting
by currents was minimal.
MIS 3, 2, and 1 are surprisingly alike in the cores. They both consist of clusters with no
prominent indication of current influence. It can evidence that the currents were slower than
before during this period, and other factors will support this implication later. Apparently,
the current during glacial MIS 2 was not as slow as during MIS 4. The small-scale changes
47
in strength of the flow were not pronounced in the record probably due to a lower sampling
resolution in the upper parts of the cores. Core 13 does not show many similarities with other
cores regarding MIS 1, implying that this layer can be even thinner and not seen at this
sampling resolution.
The conclusion can be made that cluster analysis gives good results on the identification of
features of each period when working with high-resolution cores. The results support the
hypothesis of lack of deposition in Core 13 from the second half of MIS 5 to MIS 2. The
precise timing of the beginning of MIS 1 has not been determined though.
5.7. Reconstruction of the environment and activity of currents
5.7.1. Dynamics of the position of the current and its influence on sedimentation
The change in position of the current in a channel’s profile and its influence on the
sedimentation mode can be investigated by imaging sediments layers and a current’s
strength, using the combination of echo-sounding data, preliminary age model and SS record
(Fig. 24). Based on a shape of profile (symmetry and sedimentation rate) and supposed
strength of a current, the position of the strongest current’s vector was determined. Firstly,
the choice of the direction of the flow should be explained. There are two quite solid
arguments for the flow direction from north-west to south-east. One is the shape of the
channel looking from above (Fig. 23): it is narrow on the north-west and wide on the southeast what corresponds to the expected morphology when the energy dissipates during the
movement causing sediments deposition downstream, what happens with submarine
canyons and channels (Puig et al., 2014). The evidence for the second argument is seen in
the profile: earlier the profile was relatively symmetric, later the sedimentation on the southwestern part slowed down and even reached zero while north-eastern slope gained a much
greater amount of sediments. This indicates movement of a center of the current in
northeastern direction. It has been proven that submarine channels are subject to Coriolis
force (Cossu et al., 2015) meaning movement of the vector of a flow to the right from the
direction in Northern hemisphere and to the left in Southern hemisphere. Applying this
principle to the case of our study gives the result of southeastern directed flow (Fig. 23).
This proves an assumption of Woodgate et al., 2001 about an intermediate depth current
that detaches from a northward flowing current to the west of the Ridge, turns to the southeast and moves to the Makarov Basin.
48
13, 12, 11
Figure 23. Bathymetry chart of the area (Stein, 2019). Black dot indicates
position of the cores, gray arrow – supposed direction of the current.
Changes in the current’s position can explain paradoxical sedimentation rates. The currents
were the strongest at the beginning of the period (MIS 7-6), yet the sedimentation rates were
very high. Later the current became weaker accompanying surprisingly low rates of
deposition. The reconstructed picture (Fig. 24) demonstrates that earlier the profile was quite
flat and the channel was wide, supposedly evidencing that the current was constantly
changing its position so that the energy was equally distributed on the bottom, leading to
high uniform sedimentation. After MIS 6 the current concentrated in the south-western part,
probably meeting a rocky slope to the right of its direction, thus narrowing channel down.
The grain-size data supports this as the evidence of winnowing (coarsening) was observed
on the bottom of the channel while the flank experienced accumulation of finer particles,
supposedly redeposited from the center of the channel. Later, even though the current’s
speed was not great, the deposition could not be fast due to high energy concentration.
Apparently, during MIS 1 the current started to meander again (probably because it became
shallower), leading to a thin sediment cover all over the channel. Such behavior is a common
feature of the submarine channels (Sylvester et al., 2011). Apparently, a downcore grain49
Figure 24. Evolution of the morphology and sedimentation of channel, position and
strength of the current during the end of MIS 7 – present time. The dots indicate position
of the cores: red – 11, green – 12, blue – 13. Circles with dots show the currents’ direction
and strength.
50
size fluctuations can reveal not only a temporal change in strength of a current but the
position of a current in a profile (Revel et al., 1996).
Based on that, the dynamics of the position of the current can be summarized. There is
evidence that it has gradually slowed down in the studied area, and it happened more actively
on the slope than in the middle of the channel. Apart from that, it has been moving in the
south-western direction. An implication can be made that the flow will elevate again and
spear over the channel, causing uniform deposition.
5.7.2 Changes of the sedimentological environment during the covered period
The results of numerous analyses allow reconstructing the environment and conditions of
sedimentation in the channel during the covered period i.e., from the second half of MIS 7
to the present day. In order to create a generalized simplified picture for the area, average of
the parameters in cores 11 and 12 was plotted for every MIS along with the data on core 13.
The parameters demonstrating major differences between MIS are represented in Fig. 25 AC.
TOC does not show any periodic variations; the content is quite low during the whole period,
as expected in the Arctic Ocean with relatively low production (Stein and Macdonald, 2004).
Two prominent peaks are seen: during MIS 6, excluding core 13, and MIS 1. The first one
reflects expected TOC content during the cold period, as the environmental conditions were
favorable for organic matter preservation (Stein, 2008). The bottom of the channel could
have been already under the influence of the intense current at that time, leading to reduced
biota conservation. The peak on the surface is probably caused by high biomass of living
organisms which have not been reworked.
SS record reflects consecutive alternations of high and low numbers during every MIS.
Stronger current in an interglacial period was always followed by a deceleration during
glacial. It coincides with the results of several paleocurrent studies which supposed a
connection between current speed and glaciations: there was evidence that the currents are
stronger during interglacials (Bianchi and McCave, 1999; Knutz, 2008; McCave and
Andrews, 2019).
Moreover, the general decreasing in time trend is seen. Nevertheless, in Core 13 the current
was apparently the strongest during MIS 4-2, as the flow of water with particles went by
without any deposition or even eroding the bottom.
51
Sand record recalls that of SS, although
a gradual decrease is observed during
MIS 4-2, while there is a peak of SS in
A
MIS 3. Similar character of distribution
in earlier periods can indicate currentsorted nature of this fraction, especially
considering that at that time water flow
was the strongest. A difference between
sand distribution and SS later evidences
about diverse nature of their deposition,
B
thus supporting IRD presence in MIS 4
as it was supposed earlier. Moreover,
change in a current’s mode explains a
lower amount of coarse fraction on the
flank of a channel: as the flow
concentrated in the south-east after MIS
6, the winnowing became weaker on the
northeastern flank while the deposition
of the fine fractions, lifted by the flow
C
from the bottom of the channel,
enhanced.
Figure 25. Bar-charts of TOC (A), 𝑆𝑆 (B) and sand
content (C) averaged for cores 11 and 12 and in Core
13. Arrows indicate major changes between MIS.
52
5.7.3. Water composition and climate as main factors of circulation
As it was emphasized in Chapter 2, the Arctic Ocean is a semi-enclosed basin strongly
depending on external input of water, particularly from the northern Atlantic Ocean. Apart
from that, the formation of sea ice along the continental margins leads to a brine water
formation (Shapiro et al., 2003). Such composition makes it possible to distinguish these two
types of water and change of the balance between them in the past. The study of Haley et
al., 2008 uses the ratio of neodymium isotopes from the metal-oxide coatings of the sediment
in order to reconstruct the composition of Arctic Intermediate Water. The reconstruction
showed considerably high Atlantic Water input during warm periods, and domination of
brines input during glacials MIS 6 and 4 (Fig. 26). This variability coincides with our SS
records indicating stronger flows during warm periods, making the connection between
Atlantic Water input and the speed of currents clear. Even the relatively high SS during MIS
2 coincides with unexpectedly low brine formation. The increased role of the Atlantic water
during this period is explained by the lack of ice sheet on the Kara Sea shelf region after 50
kyr (Svendsen et al., 2004), what led to the reduced brine input.
Figure 26. Nd isotopes ratio during last 250 ka; blue areas indicate
periods of low Atlantic water input and coincide with MIS 6 and MIS 4
(Haley et al., 2008, modified by R. Spielhagen, 2017).
Climate changes are the reason for the periodicity of water composition: low temperatures
cause extensive glaciations which create a thick layer of cold brine water on the surface,
thus, deepening halocline and inhibiting stratification, causing the saline Atlantic water
flows to sink and preventing their penetration far to the East (McCave and Hall, 2006).
Moreover, the formation of Atlantic Intermediate Water happened further south in the North
Atlantic (Ganopolski et al., 1998), limiting the influence on the Arctic Ocean. During
warmer times the cold halocline was thinner and closer to the surface, the stratification was
enhanced and the intermediate water from Atlantic formed close to the north and reached
further east (Haley et al., 2008; Hoffman, 2019; Jakobsson et al., 2014; Knutz, 2008; Stein,
2008).
53
It has been known that the Lomonosov Ridge is the largest obstacle for the water flowing
from the Atlantic to the east, dividing the ocean into two large basins and creating a
difference in water composition. Some part of the water reaches Amerasian Basin as contour
currents along the Eurasian shelf bypassing the ridge. Yet, there was evidence that
intermediate layers go through the ridge (see Chapter 1). This study reveals that the channel
with water depth only about 1700 m can be used for such purpose.
6. Conclusions and outlook
The paleo-current activity was reconstructed in a channel in the Southern Lomonosov Ridge
near the Eurasian shelf at 1700 m water depth covering the time period from the end of MIS
7 to present days. Grain-size and total organic carbon data from three gravity-cores were
used. Quite low TOC contents, characteristic for this area, were revealed. Dominating
sediments were very fine in grain-size: clayey silt and silty clay, yet numerous layers with
strong sand presence were discovered.
Sortable silt mean size was the main proxy applied for the reconstruction of the currents’
speed. A strong influence of currents on the sedimentation was indicated in some parts of all
three studied cores. The attempt to distinguish between IRD and coarse particles transported
by currents was made; weak positive correlation of the sand fraction with the sortable silt
mean supposed that the currents were strong enough to transport sand; therefore, this fraction
consisted from IRD only partially. Yet, several peaks of IRD input were detected during the
terminations.
The sedimentation pattern demonstrated features of a contourite system: lower rates of
deposition at the bottom of a moat and accumulation at the flank. The combination of the
age model, sedimentation rates and grain-size data allowed to reconstruct the dynamics of
the intermediate water flow. Both temporal and spatial changes in current speed, as well as
the direction of the current, were revealed. Based on this reconstruction, the current flowing
from north-west to south-east during the whole period has experienced a gradual decrease in
strength. It was assumed that the flow was widely distributed over the channel during MIS
7-6 leading to a high sedimentation rate all over the cores sites, then it became reduced and
concentrated along the south-western slope where the increase in energy caused extremely
low deposition and erosion together with a coarsening. The north-eastern flank was subject
to an active deposition of fine-grained sediments, transported by the current.
Periods of stronger flow generally coincided with the interglacial stages and terminations. It
was explained by an active Atlantic Water inflow into the Arctic Ocean during interglacial
54
stages, when several factors influenced the formation of water masses, such as the formation
of water closer to the north and stronger stratification allowing for the intermediate water to
reach further east. This evidences that water from the Atlantic Ocean could not only go into
the Makarov Basin as a contour current along the shelf slope but also cross the Lomonosov
Ridge through the channel.
Therefore, this study can be considered as a successful application of the sortable silt mean
approach as a current speed proxy in the Arctic Ocean. Nevertheless, some aspects can be
improved and developed. In order to precisely distinguish IRD in the sediments, a more
detailed study of the coarse fraction should be performed, such as mineralogical and grain
shape analyses. For a more precise determination of the timing of water inflows and currentcontrolled processes, the existing age model has to be improved significantly. In this context,
the period of special interest is the last interglacial MIS 5 with its sub-stages which did not
receive close inspection in the current study. Furthermore, for reconstructing the long-term
history and evolution of the channel during Cenozoic times, seismic data could be used.
55
Acknowledgements
I would like to thank Saint Petersburg State University, Hamburg University and University
of Bremen, Federal Ministry of Education and Research, project The Changing Arctic
Transpolar System (CATS), 03F0776, and Polar and Marine Sciences master program
POMOR for the opportunity to write this master thesis using the knowledge acquired both
in Russia and Germany.
I thank my supervisor from German side Prof. Dr. Rüdiger Stein (AWI) for an invitation to
the expedition Polarstern 115/2 as well as useful comments and advise on the work, and, of
course, the idea. I thank my supervisor from Russian side Dr. Alexey Krylov (SPSU) for the
valuable remarks and questions.
I also thank captain and crew of RV Polarstern together with scientific group onboard,
especially the geology party, for the assistance with the fieldwork.
Moreover, I am grateful to:
-
AWI for providing me with access to the laboratories and help of the technicians
(Valéa Schumacher and others);
-
Dr. Frank Niessen (AWI) for the echo-sounding and physical properties data;
-
Sergei Freiman (Moscow State University) for the fruitful discussion on the topic.
56
References
Aagaard, K., Swift, J. H., and Carmack, E. C., 1985. Thermohaline circulation in the Arctic
Mediterranean Seas. Journal of geophysical research, 90: 4833–4846.
Andrews, J.T., 2011. Unravelling sediment transport along glaciated margins (the
northwestern Nordic Seas) using quantitative X-ray diffraction of bulk (< 2 mm)
sediment. In: Bhuiyan, A.F. (Ed.), Sediment Transport. InTech, 225–248.
Andrews, J.T., Klein, A.J., Jenner, K.A., Jennings, A.E., and Campbell, C., 2018. The
variability of Baffin Bay seafloor sediment mineralogy: the identification of discrete
glacial sediment sources and application to Late Quaternary downcore analysis. Can. J.
Earth Sci., 55: 620–639.
Bianchi, G. G. and McCave, I. N., 1999. Holocene periodicity in North Atlantic climate and
deep-ocean flow south of Iceland. Nature, 397: 515–517.
Bischof, J., Clark, D.L., and Vincent, J.-S., 1996. Origin of ice-rafted debris: Pleistocene
paleoceanography in the Western Arctic Ocean. Paleoceanography, 11 (6): 743–756.
Bluhm, B. A., Kosobokova, K. N., and Carmack, E. C., 2015. A tale of two basins: An
integrated physical and biological perspective of the deep Arctic Ocean. Progress in
Oceanography, 139: 89–121.
Chaddock, R. E., 1925. Principles and Methods of Statistics. Boston, Houghton Mifflin
Company, 471 p.
Clark, D.L., Whitman, R.R., Morgan, K.A., and Mackay, S.D., 1980. Stratigraphy and
glacial-marine sediments of the Amerasian Basin, central Arctic Ocean. Spec. Pap. —
Geol. Soc. Am. 181: 1–57.
Cochran, J. R., Edwards, M. H., and Coakley, B. J., 2006. Morphology and structure of the
Lomonosov Ridge, Arctic Ocean. Geochem. Geophys. Geosyst., 7: 26 p.
Coope, G.R., Lemdahl, G., Lowe, J.J. and Walkling, A.,1999. Temperature gradients in
northern Europe during the last glacial – Holocene transition (14–9
14
C kyr BP)
interpreted from coleopteran assemblages. Journal of Quaternary Science, 13(5): 419–
433.
Cossu, R., Wells, M. G., and Peakall, J., 2015. Latitudinal variations in submarine channel
sedimentation patterns: the role of Coriolis forces. Journal of the Geological Society,
172(2): 161–174.
57
Darby, D.A., Bischof, J.F., and Jones, G.A., 1997. Radiocarbon chronology of depositional
regimes in the western Arctic Ocean. Deep-Sea Research II 44 (8): 1745–1757.
ELTRA Elemental Analyzers, 2013. Carbon / Sulfur Analyzers CS-800│CS-2000, brochure
99000-0001/E-05-2013, Eltra GmbH, Germany
Faugères, J. C. and Mulder, T., 2011. Contour currents and contourite drifts. Developments
in Sedimentology, 63(C): 149–214.
Folk, R.L. and Ward, W.C., 1957. Brazos River bar: a study in the significance of grain size
parameters. Journal of Sedimentary Petrology, 27: 3–26.
Ganopolski, A., Rahmstorf, S., Petoukhov, V., and Claussen, M., 1998. Simulation of
modern and glacial climates with a coupled global model of intermediate complexity.
Nature, 391: 351–356.
Haley, B. A., Frank, M., Spielhagen, R. F., and Eisenhauer, A., 2008. Influence of brine
formation on Arctic Ocean circulation over the past 15 million years. Nature
Geoscience, 1(1): 68–72.
Hass, H. C., 2002. A method to reduce the influence of ice-rafted debris on a grain size
record from northern Fram Strait, Arctic Ocean. Polar Research 21(2): 299–306.
Henrich, R., Kassens, H., Vogelsang, E., and Thiede, J., 1989. Sedimentary facies of glacialinterglacial cycles in the Norwegian Sea during the last 350 ka. Marine Geology, 86:
283–319.
Hjulstrom, F., 1935. Studies of Morphological Activity of Rivers as illustrated by the River
Fyris. Bulletin of the Geological Institute University of Uppsala, 25: 221–527.
Hoffmann, S. S., Dalsing, R. E., and Murphy, S. C., 2019. Sortable silt records of
intermediate-depth circulation and sedimentation in the Southwest Labrador Sea since
the Last Glacial Maximum. Quaternary Science Reviews, 206: 99–110.
Ikeda, M., Tanaka, S. S., and Watanabe, Y. W., 2018. Circulation patterns in the lower Arctic
Ocean derived from geochemical data. Journal of Oceanography, 74(5): 453–470.
Jakobsson, M., 2002. Hypsometry and volume of the Arctic Ocean and its constituent seas.
Geochem. Geophys. Geosyst., 3: 1–18.
Jakobsson, M., Grantz, A., Kristoffersen, Y., and Macnab, R., 2003. Physiographic
provinces of the Arctic Ocean seafloor. GSA Bull., 115 (11): 1443–1455.
58
Jakobsson, M., Andreassen, K., Bjarnadóttir, L. R., Dove, D., Dowdeswell, J. A., England,
J. H., Funder, S., Kelly Hogan, K., Ingólfsson, Ó., Jennings, A. E., Larsen, N. K.,
Kirchner, N., Landvik, J. Y., Mayer, L. A., Mikkelsen, N., Möller, P., Niessen, F.,
Nilsson, J., O’Regan, M., Polyak, L., Nørgaard-Pedersen, N., and Stein, R., 2014.
Arctic Ocean glacial history. Quaternary Science Reviews, 92: 40–67.
Jakobsson, M., Backman, J., Rudels, B., Nycander, J., Frank, M., Mayer, and L., Moran, K.,
2007. The early Miocene onset of a ventilated circulation regime in the Arctic Ocean.
Nature, 447: 986–989.
Jennings, A. E., Hald, M., Smith, M., and Andrews, J. T., 2006. Freshwater forcing from the
Greenland Ice Sheet during the Younger Dryas: evidence from southeastern Greenland
shelf cores. Quaternary Science Reviews, 25: 282–298.
Jones, E. P., 2001. Circulation in the Arctic Ocean. Polar Research 20(2): 139–146.
Jonkers, L., Barker, S., Hall, I. R., and Prins, M. A., 2015. Correcting for the influence of
ice-rafted detritus on grain size-based paleocurrent speed estimates. Paleoceanography,
30: 1347–1357.
Knutz, P. C., 2008. Chapter 24. Palaeoceanographic Significance of Contourite Drifts. In:
Rebesco, M., Camerlenghi, A. (eds.), Developments in Sedimentology, Contourites,
60(08). Elsevier, 511–535.
Lamy, F., Arz, H. W., Kilian, R., Lange, C. B., Lembke-Jene, L., Wengler, M., Kaiser, J.,
Wenglera, M., Kaiserb, J., Baeza-Urreac, O., Hallf, I. R., Haradag, N., and Tiedemann,
R., 2015. Glacial reduction and millennial-scale variations in Drake Passage
throughflow. PNAS, 112(44): 13496–13501.
Lantzsch, H., Hanebuth, T. J. J., Horry, J., Grave, M., Rebesco, M., and Schwenk, T., 2017.
Deglacial to Holocene history of ice-sheet retreat and bottom current strength on the
western Barents Sea shelf. Quaternary Science Reviews, 173: 40–57.
Laukert, G., Frank, M., Bauch, D., Hathorne, E. C., Rabe, B., Appen, W. Von, and Zieringer,
M., 2017. Ocean circulation and freshwater pathways in the Arctic Mediterranean based
on a combined Nd isotope, REE and oxygen isotope section across Fram Strait.
Geochimica et Cosmochimica Acta, 202: 285–309.
Ledbetter, M. T., 1986. A late Pleistocene time-series of bottom-current speed in the Vema
Channel. Palaeogeogr. Palaeoclimatol. Palaeoecol., 53: 97–105.
59
Lewis, A.R., Marchant, D.R., Ashworth, A.C., Hemming, S.R., and Machlus, M.L., 2007.
Major middle Miocene global climate change: Evidence from East Antarctica and the
Transantarctic Mountains. GSA Bulletin, 119 (11-12): 1449–1461.
Lisiecki, L.E. and Raymo, M.E., 2005. A Pliocene-Pleistocene stack of 57 globally
distributed benthic d18O records. Paleoceanography, 20: 17 p.
Li, N., Yang, X., Peng, J., Zhou, Q., and Su, Z., 2019. Deep-water bottom current evolution
in the northern South China Sea during the last 150kyr: Evidence from sortable-silt
grain size and sedimentary magnetic fabric. Journal of Asian Earth Sciences, 171: 78–
87.
Lototskaya A. and Ganssen G. M., 1999. The structure of Termination II (penultimate
deglaciation and Eemian) in the North Atlantic. Quaternary Science Reviews, 18 (14):
1641–1654.
Mora, C.I., Driese, S.G., and Colarusso, L. A., 1996. Middle to Late Paleozoic Atmospheric
CO2 Levels from Soil Carbonate and Organic Matter. Science, 271 (5252): 1105.
Revel M., Cremer M., and Grousset F.E., 1996. Grain-size and Sr-Nd isotopes as tracer of
paleo-bottom current strength, Northeast Atlantic Ocean. Marine Geology, 13: 233–
249.
Mao, L., Piper, D. J. W., Saint-Ange, F., and Andrews, J. T., 2018. Labrador Current
fluctuation during the last glacial cycle. Marine Geology, 395: 234–246.
Mauritzen C., Rudels B., and Toole J., 2013. The Arctic and Subarctic oceans/seas. In:
Siedler G., Griffies S.M., Gould J, and Church J.A. (eds.), Ocean circulation and
climate. Amsterdam: Academic Press, 443–470.
McCave, I. N., Manighetti, B., and Robinson, S. G., 1995. Sortable Silt and Fine Sediment
Size/Composition
Slicing:
Parameters
for
Palaeocurrent
Speed
and
Palaeoceanography. Paleoceanography, 10(3): 593–610.
McCave, I. N., 2008. Chapter 8. Size Sorting During Transport and Deposition of Fine
Sediments: Sortable Silt and Flow Speed. In: Rebesco, M., Camerlenghi, A. (eds.),
Developments in Sedimentology, Contourites, 60(08). Elsevier, 121–142.
McCave, I. N., Thornalley, D. J. R., and Hall, I. R., 2017. Relation of sortable silt grain-size
to deep-sea current speeds: Calibration of the ‘Mud Current Meter’. Deep-Sea Research
Part I: Oceanographic Research Papers, 127: 1–12.
60
McCave, I. N. and Andrews, J. T., 2019. Distinguishing current effects in sediments
delivered to the ocean by ice. I. Principles, methods and examples. Quaternary Science
Reviews, 212: 92–107.
McCave, I. N. and Hall, I. R., 2006. Size sorting in marine muds: Processes, pitfalls, and
prospects for paleoflow-speed proxies. Geochem. Geophys. Geosyst., 7: 37 p.
Moseley, G. E., Spötl, C., Cheng, H., Boch, R., Min, A., and Edwards, R. L., 2015.
Termination-II interstadial/stadial climate change recorded in two stalagmites from the
north European Alps. Quaternary Science Reviews, 127: 229–239.
Obrochta, S. P., Crowley, T. J., Channell, J. E. T., Hodell, D. A., Baker, P. A., Seki, A., and
Yokoyama, Y., 2014. Climate variability and ice-sheet dynamics during the last three
glaciations. Earth and Planetary Science Letters, 406: 198–212.
Oppo, D. W., Keigwin, L. D., and McManus, J. F., 2001. Persistent suborbital climate
variability in marine isotope. Paleoceanography, 16(3): 280–292.
Piskarev, A. L., Butsenko, V. V, Chernykh, A. A., Ivanov, M. V, Kaminsky, V. D., Poselov,
V. A., and Savin, V. A., 2018. Lomonosov Ridge. In: Piskarev A., Poselov V.,
Kaminsky V. (eds), Geologic Structures of the Arctic Basin. Springer, Cham, 157–185.
Proshutinsky, A., Dukhovskoy, D., Timmermans, M., Krishfield, R., and Bamber, J. L.,
2015. Arctic circulation regimes. Phil. Trans. R. Soc. A, 373: 18 p.
Puig, P., Palanques, A., and Martín, J., 2014. Contemporary Sediment-Transport Processes
in Submarine Canyons. Annual Review of Marine Science, 6(1): 53–77.
Rebesco, M., Hernández-Molina, F. J., Van Rooij, D., and Wåhlin, A., 2014. Contourites
and associated sediments controlled by deep-water circulation processes: State-of-theart and future considerations. Marine Geology, 352: 111–154.
Rudels, B., Jones, E. P., Anderson, L. G., and Kattner, G., 1994. On the Intermediate Depth
Waters of the Arctic Ocean. In: Johannessen, O. M., Muench, R. D., and Overland, J.
E. (eds.), The Polar Oceans and Their Role in Shaping the Global Environment.
Geophysical monograph, 85. American Geophysical Union, 33–46.
Rudels, B., 2015. Arctic Ocean circulation, processes and water masses: A description of
observations and ideas with focus on the period prior to the International Polar Year
2007 – 2009. Progress in Oceanography, 132: 22–67.
61
Rudels, B., Korhonen, M., Schauer, U., Pisarev, S., Rabe, B., and Wisotzki, A., 2015.
Circulation and transformation of Atlantic water in the Eurasian Basin and the
contribution of the Fram Strait inflow branch to the Arctic Ocean heat budget. Progress
in Oceanography, 132: 128–152.
Rudels, B., 2018. Arctic Ocean Circulation. In: Cochran, J. K., Bokuniewicz, H. J., Yager,
P.L. (eds.), Encyclopedia of Ocean Sciences (3rd ed.). Elsevier Inc. Academic Press,
262–277.
http://www.r-project.org The R Project for Statistical Computing, accessed on February 20
2019
SediGraph III PLUS Particle Size Analyzer brochure 512/42705/00 Micromeritics
Instrument Corporation, USA
Shapiro, G. I., Huthnance, J. M., and Ivanov, V. V., 2003. Dense water cascading off the
continental shelf. J. Geophys. Res., 108: 3390.
Smillie, Z., Stow, D., and Esentia, I., 2018. Deep-Sea Contourites Drifts, Erosional Features
and Bedforms. In: Cochran, J. K., Bokuniewicz, H. J., Yager, P.L. (eds.), Encyclopedia
of Ocean Sciences, (3rd ed.). Elsevier Inc. Academic Press, 97–110.
Spall, M. A., 2013. On the Circulation of Atlantic Water in the Arctic Ocean. J. Phys.
Oceanogr., 43: 2352–2371.
Spielhagen, R. F., Baumann, K. H., Erlenkeuser, H., Nowaczyk, N. R., Nørgaard-Pedersen,
N., Vogt, C., and Weiel, D., 2004. Arctic Ocean deep-sea record of northern Eurasian
ice sheet history. Quaternary Science Reviews, 23(11–13): 1455–1483.
Stein, R., 1985. Rapid grain-size analyses of clay and silt fraction by SediGraph 5000D:
Comparison with Coulter Counter and Atterberg methods. Journal Of Sedimentary
Research, 55(4): 590–593.
Stein, R., 2008. Arctic Ocean Sediments: Processes, Proxies, and Paleoenvironment. Oxford,
Elsevier, 602 p.
Stein, R. (Ed.), 2015. The Expedition PS87 of the Research Vessel POLARSTERN to the
Arctic Ocean in 2014. Reports on Polar and Marine Research, Bremerhaven, Alfred
Wegener Institute for Polar and Marine Research, 688: 273 p.
Stein, R., Fahl, K., Gierz, P., Niessen, F., and Lohmann, G., 2017. Arctic Ocean sea ice cover
during the penultimate glacial and the last interglacial. Nature Communications, 8: 373.
62
Stein, R. (Ed.), 2019. The Expedition PS115/2 of the Research Vessel POLARSTERN to
the Arctic Ocean in 2018. Reports on Polar and Marine Research, Bremerhaven, Alfred
Wegener Institute for Polar and Marine Research, 728: 249 p.
Stein, R. and Macdonald, R. W., 2004. Geochemical Proxies used for Organic Carbon
Source Identification in Arctic Ocean Sediments. In: Stein, R. and Macdonald, R. W.
(Ed.), The organic carbon cycle in the Arctic Ocean. Springer, 24–32.
Stow, D. A. V., Hunter, S., Wilkinson, D., and Hernández-Molina, F. J., 2008. Chapter 9.
The Nature of Contourite Deposition. In: Rebesco, M., Camerlenghi, A. (eds.),
Developments in Sedimentology Contourites, 60(08). Elsevier, 143–156.
Stroeve, J. and Notz, D., 2018. Changing state of Arctic sea ice across all seasons. Environ.
Res. Lett., 13: 24 p.
Svendsen, J.I., Alexanderson, H., Astakhov, V.I., Demidov, I., Dowdeswell, J.A., Funder,
S., Gataullin, V., Henriksen, M., Hjort, C., Houmark-Nielsen, M., Hubberten, H.W.,
Ingølfsson, Ó., Jakobsson, M., Kjær, K.H., Larsen, E., Lokrantz, H., Lunkka, J.P., Lyså,
A., Mangerud, J., Matiouchkov, A., Murray, A., Möller, P., Niessen, F., Nikolskaya,
O., Polyak, L., Saarnisto, M., Siegert, C., Siegert, M.J., Spielhagen, R.F., and Stein, R.,
2004. Late quaternary ice sheet history of northern Eurasia. Quaternary Science
Reviews, 23 (11–13): 1229–1271.
Swift J.H. and Koltermann K.P., 1988. The origin of Norwegian Sea deep water. J. Geophys.
Res.: Oceans, 93 (C4): 3563–3569.
Sylvester, Z., Pirmez, C., and Cantelli, A., 2011. A model of submarine channel-levee
evolution based on channel trajectories: Implications for stratigraphic architecture.
Marine and Petroleum Geology, 28(3): 716–727.
Thomas, A. L., Henderson, G. M., and McCave, I. N., 2006. Constant flow of AABW into
the Indian Ocean over the past 140 ka? Conflict between
231
Pa/230Th and sortable silt
records. Goldschmidt Conference Abstracts: 647.
Wang, R., Polyak, L., Xiao, W., Wu, L., Zhang, T., and Sun, Y., 2018. Late-Middle
Quaternary lithostratigraphy and sedimentation patterns on the Alpha Ridge, central
Arctic Ocean: Implications for Arctic climate variability on orbital time scales.
Quaternary Science Reviews, 181: 93–108.
63
Woodgate, R., 2012. Arctic Ocean Circulation: Going Around At the Top Of the World.
Nature Education Knowledge, 4(8): 1–13.
Woodgate, R. A., Aagaard, K., Muench, R. D., Gunn, J., Rudels, B., Roach, A. T., and
Schauer, U., 2001. The Arctic Ocean Boundary Current along the Eurasian slope and
the adjacent Lomonosov Ridge: Water mass properties, transports and transformations
from moored instruments. Deep-Sea Research, I 48: 1757–1792.
64
Appendix
The appendix is written on a CD enclosed to the thesis. The content is as follows:
- core photographs (Author: R. Love);
- Excel file with data tables on TOC, grain-size fractions content in % (sand, very coarse
silt, coarse silt, medium silt, fine silt, very fine silt, and clay) and calculated parameters
(SS, SS%, SS in a phi scale, SSpot, delta SS, and sorting/standard deviation) for cores 11, 12,
and 13.
65
Statement on the thesis’ originality
Herewith I, Elena Popova, declare that I wrote the thesis independently and did not
use any other resources than those named in the bibliography, and, in particular, did not use
any internet resources except for those named in the bibliography. The master thesis has not
been used previously as part of an examination. The master thesis has not been previously
published.
_________________ Elena Popova
66
Отзывы:
Авторизуйтесь, чтобы оставить отзыв