The East Siberian Arctic Shelf (ESAS) is the broadest and shallowest continental shelf in the world. It is characterized by both the highest rate of coastal erosion in the world and a large riverine input of terrigenous dissolved organic matter (DOM). DOM plays a significant role in marine aquatic ecosystems. The chromophoric fraction of DOM (CDOM) directly affects the quantity and spectral quality of available light, thereby impacting both primary production and ultraviolet (UV) exposure in aquatic ecosystems.
A multiyear study of CDOM absorption, fluorescence, and spectral characteristics was carried out over the vast ESAS in the summer–fall seasons. The paper describes observations accomplished at 286 stations and 1766 in situ high-resolution optical measurements distributed along the nearshore zone. Spatial and interannual CDOM dynamics over the ESAS were investigated, and driving factors were identified. It was shown that the atmospheric circulation regime is the dominant factor controlling CDOM distribution on the ESAS.
This paper explores the possibility of using CDOM and its spectral
parameters to identify the different biogeochemical regimes in the surveyed
area. The analysis of CDOM spectral characteristics showed that the major
part of the Laptev and East Siberian seas shelf is influenced by terrigenous
DOM carried in riverine discharge. Western and eastern provinces of the ESAS
with distinctly different DOM optical properties were also identified; a
transition between the two provinces at around 165–170
In the western ESAS, a region of substantial river impact, the content of aromatic carbon within DOM remains almost constant. In the eastern ESAS, a gradual decrease in aromaticity percentage was observed, indicating contribution of Pacific-origin waters, where allochthonous DOM with predominantly aliphatic character and much smaller absorption capacity predominates. In addition, we found a stable tendency towards reduced concentrations of CDOM and dissolved lignin and an increase in spectral slope and slope ratio values eastward from the Lena River delta; the Lena is the main supplier of DOM to the eastern Arctic shelf.
The strong positive correlation (
Current climate change is particularly evident and amplified in the high
latitudes of the Northern Hemisphere. The system change is characterized by
an increase in the average annual temperature and atmospheric circulation
intensity, a reduction in sea ice coverage and thickness, accelerated
degradation of permafrost, and an increase in coastal and bottom
erosion and river flows (IPCC, 2013). Unlike other oceans, the Arctic Ocean
is completely surrounded by permafrost. The Arctic region contains an
abundance of organic carbon (OC) buried inland and within the sedimentary
basin of the Arctic Ocean, which might become a part of the current marine
biogeochemical cycle due to thawing of on-land and sub-sea permafrost,
increased coastal and bottom erosion, accelerated river discharge, and
soil-based carbon losses (Günther et al., 2013; Shakhova et al., 2009,
2017; Semiletov, 1999; Vonk and Gustafsson, 2013; Vonk et al., 2012, 2014).
The recent studies accomplished within the framework of the International
Siberian Shelf Study project (Semiletov and Gustafsson, 2009; Semiletov et
al., 2011; Tesi et al., 2014; Bröder et al., 2016; Charkin et al., 2011)
demonstrate that coastal erosion is the main source of particulate OC (POC)
to the East Siberian Arctic Shelf (ESAS), the broadest and shallowest shelf
in the world. Oxidation of POC exported from thawing coastal (and
bottom) permafrost and freshening due to growing Siberian river runoff from
extensive permafrost-underlain watersheds play a major role in the severe
ESAS acidification that has been reported (Semiletov et al., 2016). At the
same time, river discharge to shelf waters supplies terrestrial carbon in
the form of dissolved OC (DOC) and waters enriched by carbon dioxide (CO
Annually, the Arctic rivers transport 25–36 Tg of DOC to the Arctic Ocean,
which is
The absorbance and fluorescence properties of the chromophoric fraction of DOM (CDOM) are “optical markers”, comparable to traditional biomarkers used in geochemistry (e.g., lignin) (Stedmon and Nelson, 2015). According to Coble (2007), up to 70 % of DOM in shelf waters is represented by CDOM, which is critical in a number of biochemical and photochemical processes and which defines the optical properties of natural waters, thus affecting the depth of the photic layer (Granskog et al., 2007; Hill, 2008). Quantitative descriptions of the dynamics and variability in CDOM optical properties are often required, particularly in coastal waters, in order to accurately predict light penetration and primary production (Matsuoka et al., 2007). It should be noted that reliable estimation of coastal water optical characteristics is crucial for validating and calibrating remote sensing data processing results (Vantrepotte et al., 2012; Bondur and Vorobev, 2015). High CDOM values are typical for the eastern Laptev Sea; in the western East Siberian Sea they result from the large discharge of the Lena River, which is characterized by high CDOM concentrations (Semiletov et al., 2013; Pugach and Pipko, 2013). Most likely, the color that indicates the presence of CDOM in this water, when seen from space, is due to the presence of chlorophyll (Heim et al., 2014). This may explain the recently proposed increase in the net primary production rates reported for these ESAS areas based on satellite data interpretation (Arrigo and van Dijken, 2011).
Empirical relationships between the optical properties of DOM and DOC have already been the subject of investigation in the Arctic Ocean (Stedmon et al., 2011; Guéguen et al., 2005, 2007; Fichot and Benner, 2011; Gonçalves-Araujo et al., 2015; Kaiser et al., 2017). The present study synthesizes the authors' multiyear observations on the remote ESAS with the focus on exploring the extent and dynamics of riverine DOC, using CDOM as a proxy.
This paper aims (1) to study the spatial and interannual dynamics of DOM optical characteristics in shelf waters of the eastern Arctic seas on the basis of multiyear summertime (August –September) expedition data (2003, 2004, 2005, 2008, 2011); (2) to examine the relationship between CDOM fluorescence and DOC in order to validate a useful method for accurately predicting DOC concentration from CDOM properties in the ESAS; and (3) to demonstrate the feasibility, using DOM optical characteristics, of determining the terrigenous DOM distribution and identifying different biogeochemical provinces in the shelf water.
Study area: location of oceanographic stations at which measurements were taken in August–October 2003, 2004, 2005, 2008, and 2011. Location of selected stations on the ESAS in September 2004 is shown in color.
This study is focused on the Laptev and East Siberian seas, where the
influence of river discharge and the biogeochemical signal of permafrost
degradation are the most prominent (Alling et al., 2010; Nicolsky and
Shakhova, 2010; Sánchez-Garcia et al., 2011; Shakhova et al., 2015,
2017; Vonk et al., 2012). The shelf area of these seas covers 40 % of the
Arctic shelf and 20 % of the entire Arctic Ocean area (Stein and
Macdonald, 2004). Annually, 767 km
The East Siberian Sea is one of the least-studied seas; this lack of
attention is mainly attributed to severe icebound conditions. The Kolyma
and Indigirka are the main rivers for which the East Siberian Sea serves as
a receiving basin; however, their total average annual discharge does not
exceed 180 km
This paper presents data from research expeditions conducted during the
summer–fall seasons of 2003 and 2004 (HV
The study was carried out using a hydrological
conductivity–temperature–depth (CTD) SBE 19plus SeaCAT Profiler probe
(
Dates, number of stations, samples collected, and parameters measured in the surface layer during cruises.
In 2003, 2004, 2005, and 2011 the measurements were carried out at
286 oceanographic stations and in 2008 along the ship's track using a seawater
intake (SWI) system. Water was pumped from 4 m depth at 30 L min
In 2004, DOC concentration was measured in an International Arctic Research
Center laboratory at the University of Alaska (Fairbanks, Alaska, USA);
in 2008, DOC concentration was measured
onboard the RV
Absorbance of CDOM was measured using a UNICO 2804 spectrophotometer with a
1 cm quartz cuvette over the spectral range from 200 to 600 nm at 1 nm
intervals (Table 1). Milli-Q (Millipore) water was used as the reference for
all samples. Water samples underwent filtration through acid-washed Whatman
glass fiber filters (GF/F, nominal pore size 0.7
The absorption coefficient (
The absorption coefficients at 350 nm (
The dependence of
The spectral slope,
Specific UV absorbance (SUVA), defined as the UV absorbance of a water sample at
254 nm normalized for DOC concentration, is used to estimate the degree of
aromaticity in bulk CDOM (Weishaar et al., 2003):
This equation is applicable for a wide range of aquatic environments (seas,
bogs, lakes) since the authors used humic substances that have different
chemical characteristics and demonstrated a strong correlation (
Dissolved lignin is a well-established biomarker of terrigenous DOM in the
ocean and has been successfully applied as a tracer of riverine inputs in
the Arctic Ocean (Fichot et al., 2016). To assess its content in coastal
waters, two empirical models for the retrieval of the sum of nine lignin
phenols (TDLP
Data were tested statistically using an empirical distribution function test
in the Statistics 7.0 software package. Descriptive statistics were
calculated for the 95 % confidence interval of the mean (
The spatial distribution of sea surface salinity obtained in September 2003, 2004, 2005, 2008, and 2011 is shown in Fig. 2a. During summertime, the coastal currents transport a large part of the Lena River water eastward into the East Siberian Sea. The overall hydrological conditions in the nearshore zone were mainly determined by the interaction of river and marine water (Fig. 2a). The sea surface salinity showed a general eastward increasing trend. The salinity values varied between 3.1 (2004) and 32.34 (2011), with the lowest values associated with freshwater input from the Lena River and the higher values attributed to the presence of Pacific water. The maximum eastward spreading of river water was detected in September 2004 when the freshwater signal was found in the vicinity of Long Strait.
Comparative analysis of 2003, 2004, 2005, 2008, and 2011 data showed that, in general, CDOM concentration in ESAS surface waters varies depending upon distance from the river water source: the maximum CDOM was recorded near river mouths, and the minimum was recorded in regions remote from direct river discharge. However, the spatial distribution of waters with high CDOM values on the ESAS differed significantly in different years (Fig. 2b). The CDOM isoline equal to 15 QSU in Fig. 2b can be used to compare the distribution of river water in different years. Location of this CDOM isoline correlates spatially with isohaline 24.5, which has been suggested as a marker for the boundary of surface shelf waters diluted by riverine runoff in the Siberian seas (Nikiforov and Shpaikher, 1980; Semiletov et al., 2000, 2005).
Comparison of the discharge of the main rivers flowing into the Laptev and
East Siberian seas (the Lena and Kolyma rivers), which was based on Tiksi
Hydromet data and the
Following Proshutinsky et al. (2015), dominant atmospheric processes and differences in ice-coverage area mainly determine surface hydrology in the wind-driven ESAS. Therefore, we assume that interannual variability in the sea surface CDOM distribution on the ESAS is also determined by atmospheric processes.
National Centers for Environmental Prediction (NCEP) sea level pressure (SLP)
data were employed to describe the atmospheric circulation over the
Arctic Ocean (
Distributions of surface salinity
During the 2003 summer season, cyclonic atmospheric circulation
dominated over the central Arctic Ocean. SLP as low as 1005 mbar extended
over the Laptev and the East Siberian seas (Fig. 2c). The development of a
deep atmospheric depression caused onshore winds on the western periphery of
the cyclone and transferred freshened shelf waters to the east (Pipko et
al., 2008; Savel'eva et al., 2008). Waters with high CDOM concentrations
spread to 165
In 2004 the summer low pressure north of the East Siberian Sea was weaker while an anticyclone formed above the Canadian Arctic Archipelago (Fig. 2c). High river discharge, ice conditions, and offshore winds determined the maximum distribution of river waters in the ice-free East Siberian Sea (Fig. 2b).
Predominance of cyclonic atmospheric circulation over the western Laptev Sea
and anticyclonic circulation over the Beaufort Sea created significant SLP
gradients and thus strong winds in 2005 (average wind speed
of up to
In 2008, the summer atmospheric pressure field was conditioned
by the dominant anticyclone over the Beaufort Sea and a weak cyclone over
Siberia, which caused southeastern winds over the ESAS (Fig. 2c). Thus,
although the river discharge was great (the Lena River maximum discharge
occurred during the investigated period), freshened waters were located in
the southeastern Laptev Sea and the western East Siberian Sea and did not
penetrate into the eastern part of the sea. It should be noted that the ice
extent was less in September 2008 compared with the previous study periods.
The whole shelf was ice-free that year and the ice boundary moved north of
80
An area of high SLP over the Arctic Basin and low SLP above the continent
occurred in the summer season (July–August) of 2011 (Fig. 2c).
The averaged meridian wind speed was 1.5 m s
The relationship between the salinity and CDOM (QSU) in the inner and middle ESAS
surface water in August–September 2003, 2004, 2005, 2008, and 2011 (blue
circles: salinity
Taken together, field data analyses showed that the prevailing type of
atmospheric circulation and the position of action centers relative to each
other were the dominant factors controlling the ESAS surface CDOM spatial
distribution, while interannual variabilities in river discharge and ice
extent were also significant. The greatest offshore propagation of waters
with high CDOM content was recorded in 2004 when a cyclone was located above
the central region of the Arctic Ocean basin and a high atmospheric pressure
field formed over the Beaufort Sea. In September 2004 the East Siberian Sea
surface salinity from Kolyma Bay (near 160
Depth profiles of salinity
A significant negative correlation between CDOM and salinity in surface ESAS
water (
Spectral characteristics (
CDOM absorption spectra for September 2004 surface waters at two stations
(stations 118 and 97) located in contrasting shelf zones (Fig. 1) are shown
in Fig. 5. Spectrum for station 60 located in the East Siberian Sea (in a
moderate zone of mixing river and ocean waters) is also presented. CDOM
absorption spectra measured for different waters have substantially
different levels of absorption. Despite CDOM dominating the absorption
spectra in the blue and UV wavebands, the absorption coefficient
The relationship between river water
The multiyear surface water
Surface CDOM absorption coefficient spectra
The
The relationship between CDOM absorption coefficient at the 350 nm
wavelength (
Figure 6a demonstrates the spatial distribution of aromatic carbon (
As was shown in previous studies, the concentrations of CDOM vary significantly within the ESAS (Belzile et al., 2006; Gonçalves-Araujo et al., 2015; Stedmon et al., 2011; Walker et al., 2013), mainly because of variations in the sources of DOM. The current multiyear investigations of the ESAS extend previous studies showing not only the inhomogeneity of spatial DOM variability but also temporal DOM variability and its spectral characteristics distribution in the shallow shelf waters.
The relationship between the dissolved lignin concentrations (TDLP
We found a stable tendency to a reduction in the concentrations of CDOM and
aromaticity and an increase in
Correlation between CDOM (QSU) and the absorption coefficient at 350 nm
(
The data obtained allowed us to calculate the lignin content in surface waters
of the inner and middle ESAS based on CDOM spectral characteristics (Fig. 7)
and to estimate the average values in each of the provinces. The most
significant difference in the average values that characterize the western
and eastern shelf regimes were found for the dissolved lignin
concentration (Table 2). Lignin is a well-established biomarker of
terrigenous OM in the ocean (Hedges et al., 1997; Fichot and Benner, 2014;
Fichot et al., 2016). It is exclusively produced on land by vascular plants.
As a result, lignin extracted from seawater and marine sediments has long
been used to derive qualitative and quantitative information about the
origins, transformations, and fates of terrigenous OM in the ocean (Amon et
al., 2012; Fichot and Benner, 2014; Tesi et al., 2014; Fichot et al., 2016).
The three large Siberian rivers, Lena, Yenisey, and Ob, which also have the
highest proportion of forests within their watersheds among the six largest
Arctic rivers, contribute about 90 % of the total lignin discharge to the
Arctic Ocean; the Lena River alone contributes about 48 % of the total
annual lignin discharge into the Arctic Ocean (Amon et al., 2012). The
calculated dissolved lignin concentration on the west ESAS in the late
summer–fall season ranged from 11.9 to 263.2 nmol L
In eastern ESAS surface waters (Pacific-influenced province) the average
dissolved lignin concentration decreased by more than an order of magnitude
compared to the western part (from 89.5 to 5.1 nmol L
The presented values of CDOM spectral parameters are comparable to findings from other studies previously conducted in this region (Stedmon et al., 2011; Walker et al., 2013; Gonçalves-Araujo et al., 2015; Mann et al., 2012, 2016). However, our data allow us to trace the dynamics of CDOM spectral characteristics in a highly dynamic shallow region of the Arctic shelf, between heterotrophic riverine waters (Stedmon et al., 2011; Walker et al., 2013; Gonçalves-Araujo et al., 2015; Mann et al., 2012, 2016) and autochthonous seawater (Fichot et al., 2016; Kaiser et al., 2017).
Conventional methods for the analysis of DOC are restricted to measurements of discrete samples and are limited to providing synoptic coverage on relatively small spatial scales. Estimating DOC concentrations via measuring optical DOM properties (absorption and fluorescence) therefore represents a compelling alternative (Fichot and Benner, 2011).
Methods to predict DOC concentrations from absorbance characteristics have been attempted since the early 1970s (Banoub, 1973; Lewis and Tyburczy, 1974). However, a prerequisite for successfully predicting DOC concentration is that the nonabsorbent DOC is at a constant or low level (Ferrari and Dowell, 1998). The processes responsible for DOC and CDOM distribution in the open ocean are typically independent, and the two pools usually demonstrate a negative rather than a positive correlation (Coble, 2007). The situation is different in shelf areas where terrestrial discharge is strong and distribution of river water and seawater controls the distribution of both DOC and CDOM. A novel method to accurately retrieve DOC concentrations from CDOM absorption coefficients for the salinity range of 0–37, typically encountered in river-dominated ocean margins, was successfully developed by Fichot and Benner (2011).
DOC concentration (
DOC distribution (
Here we present a method to estimate DOC concentration from in situ CDOM fluorescence using data collected in September 2004 and 2008 in surface waters of the shallow ESAS, where the Lena River is the main riverine source.
DOC and CDOM concentrations were measured simultaneously in the surface shelf waters of the Laptev and East Siberian seas during the 2004 and 2008 surveys to determine relationships between the two parameters.
In 2004, along with the study of filtered seawater sample absorption
coefficients, CDOM was measured in situ using a WETStar fluorometer. Strong
positive correlation (
This confirms previous results based on data obtained exclusively in the
East Siberian Sea (Belzile et al., 2006) showing that the presence of
suspended matter in water samples has little impact on CDOM values even in
regions of intensive terrigenous discharge. We have now compared the
relationship between CDOM and
The relationship between DOC and CDOM showed a strong positive correlation
for the 2004 and 2008 cruises (
An important implication of the obtained total relationship between DOC and CDOM (Fig. 9d) is the capability to constrain the DOC concentration for the ESAS surface waters in the late summer season using information about in situ CDOM fluorescence. It allows DOC values to be calculated with a high spatial resolution based on direct fluorescence measurements, avoiding various artifacts induced by filtering and storing samples and, thus, adding to the limited available data on DOC distribution over the ESAS.
Overall, this simple empirical model provides a practical means to derive
reasonably accurate estimates of DOC concentrations from CDOM in coastal
waters of the Siberian Arctic shelf (Figs. 9 and 10). The model adequately
reproduces the most important features of the measured DOC distribution in
the shelf waters. A comparison of the measured and calculated DOC indicates
that the model restored DOC values within
DOM optical characteristics, dissolved lignin, and salinity in two biogeochemical provinces of the inner and middle ESAS surface water (obtained during the 2004, 2005, and 2011 surveys).
AVG – average value; Md – median value; SD – standard
deviation.
Therefore, this rapid assessment of DOC concentrations on the ESAS using a WETStar fluorometer is an effective tool for obtaining information on DOC distribution in summer seasons. This approach is reliable over the salinity range of 3 to 24.5; the lower limit is defined as the lowest conventional salinity level at which all the major marginal filters have already been passed (Lisitsyn, 1994) and the upper limit is a border of riverine water distribution (Nikiforov and Shpaikher, 1980).
A multiyear study of DOM optical parameters and the spectral characteristics of the DOM chromophoric fraction was carried out, repeatedly probing the summer–fall seasons on the broadest and shallowest shelf in the world, the vast ESAS, from the Lena River delta in the Laptev Sea to Long Strait in the East Siberian Sea. For the first time, CDOM and DOC interannual variability in connection with atmospheric pressure fields and wind-driven water circulation was considered. The atmospheric circulation regime is the dominant factor controlling hydrography and spatial expansion of the area of freshwater influence that determines CDOM–DOM spatial distribution on the ESAS.
The dynamics of DOM optical properties provided a new insight into biogeochemical processes in the ESAS. The spectral characteristics of CDOM were applied to identify two clearly distinct biogeochemical provinces in the surveyed area. The analysis of CDOM spectral characteristics has clearly shown that the major part of the Laptev and East Siberian sea shelf is influenced by terrigenous DOM with high aromaticity and high lignin content, transported by riverine discharge.
The content of aromatic carbon within DOC remains almost constant in the western ESAS, which is the region of substantial river impact, while in the eastern ESAS, a gradual decrease in aromaticity percentage was observed, indicating an increasing contribution of Pacific-origin waters, where allochthonous DOM with predominantly aliphatic character and much lesser absorption capabilities prevails.
We found a stable tendency of CDOM and dissolved lignin concentrations to be
reduced and
The strong correlation between DOC and CDOM concentrations in surface shelf waters that are influenced by terrigenous discharge makes it possible to calculate DOC content from CDOM values assessed in situ using the WETStar fluorometer. Moreover, the reliable estimation of optical characteristics of coastal waters is crucial for validating and calibrating remote sensing data processing results. Employing optical techniques can increase the temporal and spatial coverage of DOM measurements across the ESAS and help to more accurately estimate the amount of terrigenous DOM; this estimation is necessary for understanding how carbon budgets and fluxes will be altered under future climate change scenarios.
DOM and related data are publicly and freely available at the
open-access Stockholm University Bolin Centre Database (
The authors declare that they have no conflict of interest.
This article is part of the special issue “Climate–carbon–cryosphere interactions in the East Siberian Arctic Ocean: past, present and future (TC/BG/CP/OS inter-journal SI)”. It is not associated with a conference.
This work was supported by the Russian Government (grant 14.Z50.31.0012),
the Far Eastern Branch of the Russian Academy of Sciences (FEBRAS); the
International Arctic Research Center (IARC) of the University of Alaska
Fairbanks through NOAA Cooperative Agreement NA17RJ1224; the US National
Science Foundation (nos. OPP-0327664, OPP-0230455, ARC-1023281,
ARC-0909546); and the NOAA OAR Climate Program Office (NA08OAR4600758).
Svetlana P. Pugach and Irina I. Pipko thank the Russian Foundation for Basic
Research (RFBR, no. 18-05-00559a) and Evgeny A. Shirshin and Irina V. Perminova acknowledge RFBR grant
nos. 15-05-09284 and 16-04-01753. Natalia E. Shakhova and Alexey S. Ruban
acknowledge the Russian Science Foundation (grant no. 15-17-20032).
Örjan Gustafsson acknowledges support from the Swedish Research
Council, the Knut and Alice Wallenberg Foundation, and an ERC Advanced Grant
(ERC-AdG CC-TOP project no. 695331). Igor P. Semiletov and Natalia E. Shakhova
acknowledge support from the ICE-ARC EU FP7 project. We thank Ronald Benner
for DOC measurements in water samples taken in the ESAS onboard HV