Methane Pathways in Winter Ice of Thermokarst Lakes, Lagoons and Coastal Waters in North Siberia

The thermokarst lakes of permafrost regions play a major role in the global carbon cycle. These lakes are sources of methane to the atmosphere but the methane flux is restricted by an ice cover for most of the year. The fate of methane in these waters and is poorly understood. We provide insights into the methane pathways in the winter ice cover on three different water bodies in a continuous permafrost region in Siberia. The first is a bay underlain by submarine permafrost (Tiksi Bay, TB), the second a shallow thermokarst lagoon (Polar Fox, PF) and the third a land-locked, freshwater thermokarst 5 lake (Goltsovoye Lake, GL). In total, 11 ice cores were analyzed as records of the freezing process and methane pathways during the winter season. In TB, the hydrochemical parameters indicate an open system freezing. In contrast, PF was classified as a semi-closed system, where ice growth eventually cuts off exchange between the lagoon and the ocean. The GL is a closed system without connections to other water bodies. Ice on all water bodies was mostly methane-supersaturated with respect to the atmospheric equilibrium concentration, except of three cores from the lake. Generally, the TB ice had low 10 methane concentrations (3.48–8.44 nM) compared to maximum concentrations of the PF ice (2.59-539 nM) and widely varying concentrations in the GL ice (0.02–14817 nM). Stable δCCH4 isotope signatures indicate that methane above the ice-water interface was oxidized to concentrations close to or below the calculated atmospheric equilibrium concentration in the ice of PF. We conclude that methane oxidation in ice may decrease methane concentrations during winter. Therefore, understanding seasonal effects to methane pathways in Arctic saline influenced or freshwater systems is critical to anticipate permafrost 15 carbon feedbacks in course of global warming.

. Study design on the Bykovsky Peninsula, NE Yakutia, Arctic Russia. Ice cores were recovered from three water bodies on the southern shore of Bykovsky Peninsula. Ice cores were recovered from Tiksi Bay (TB: blue), Polar Fox Lagoon (PF: red), and Goltsovoye Lake (GL: green dots). Imagery in the main map is Worldview-3 from Sep. 9, 2016 ((c) DigitalGlobe). 2011). For our study, we selected three water bodies at the southern coast of Bykovsky Peninsula: Tiksi Bay (TB), Polar Fox Lagoon (PF) and Goltsovoye Lake (GL) (Fig. 1, Tab. 1).
TB is a relatively shallow bay underlain by submarine permafrost (Overduin et al., 2016), at least close to the coast. The bay is located south-east of the Bykovskaya Channel, which is the major outlet of the Lena River through which about 25 % of the Lena's spring discharge exits the delta (Fedorova et al., 2015). The mixing of freshwater from the Lena River with saline 5 water from the Laptev Sea results in brackish conditions in the bay (Lantuit et al., 2011). TB belongs to the Buor Khaya Bay region, where the water column is usually stratified, with a colder, more saline water layer underlying the brackish surface layer (Overduin et al., 2016). The depth of the pycnocline varies between 4 and 10 m and the stratification can be disturbed by storm events. Storm surges also influence the sea level at TB, with maximum wave heights of about 1.1 m. Tidally-based sea-level oscillations have little influence on the height of storm surges and the Bykovsky Peninsula region is characterized 10 by a micro-tidal regime (Lantuit et al., 2011). PF is connected to TB by a shallow channel (Fig. 1) and therefore dominated by brackish water. PF represents a transitional stage between a land-bound thermokarst lake and a bay: a thermokarst lagoon.
Its morphology suggests that it has been formed in a thermokarst depression. The lower extent of ice-rich sediments and the 4 https://doi.org /10.5194/tc-2019-304 Preprint. Discussion started: 20 February 2020 c Author(s) 2020. CC BY 4.0 License. thermokarst lake beds lie below sea level . GL is a slightly oval-shaped thermokarst lake about 0.5 km in diameter, surrounded by Yedoma uplands at various stages of degradation. Tab. 1 lists description of the size and maximum depths of the studied water bodies. Table 1. Hydrological characteristics of water bodies of the southern Bykovsky Peninsula from which ice cores were taken for this study in spring 2017 (Strauss et al., 2018). Temperatures and electrical conductivity for GL and PF were measured in the field below the ice; temperature and salinity for TB are from Charkin et al. (2017 Goltsovoye Lake (GL) 0.46 -0.62 10 0 to 1.8 0.3 mS/cm 5 160 3 Methods 3.1 Sampling in the field 5 Ice cores were collected from TB in a transect roughly perpendicular to the shore (Fig. 1). For PF, two cores were drilled near the center of the lagoon. The sites of the five cores from GL were located along an approximately east-west transect across the lake. Tab. 1 lists the mean ice thicknesses of the sampled ice core for the locations.
The ice cores were taken with a Kovacs Mark II ice coring system (9 cm diameter), between Apr. 5 and 12, 2017. Cores were collected in triplicate from each sampling site. One core was used for temperature measurements, one was collected

Sample processing
The ice cores were processed in Potsdam from Dec. 4 to 15, 2017 (Cores 23, 24, 27, 28, 29, 30) and from the Apr. 30 to Mar. 4, 2018 (Cores 20, 21, 22). The ice cores were cut in a cold room at −15 • C with a band saw every 10 cm and stored for melting in gas-tight TEDLAR bags at 4 to 7 • C (over 1-2 days). The closed bags were evacuated with a vacuum pump before.
After melting, the bags were gently mixed and water was poured through tubing, without producing bubbles, into 100 mL 5 glass bottles for the analysis of CH 4 concentrations and δ 13 C in CH 4 . The remaining water was distributed into other sample bottles for hydrochemical measurements of pH, electrical conductivity (EC), dissolved organic carbon (DOC), δ 18 O and δD isotopes of water, as well as major anion and cation concentrations. Data points in the graphs of variation against core depth thus represent a mean value for a section of usually 10 cm within an ice core.

Hydrochemistry in ice 10
Electrical conductivity and temperature of GL and PF were measured in the field. EC and pH of ice core samples were measured with a WTW Multilab 540 device as soon as possible after bottling. The salinity was calculated from the values of the electrical conductivity after McDougall and Barker (2011). The samples for DOC were filtered with 0.7 µm pore size glass fiber filters (the filters were first rinsed with 20 mL of the sample), filled in 20 mL glass-headspace vials and closed with aluminum crimp caps. For preservation, 50 µL of 30 % supra-pure HCl were added to the sample before closing the vials, which were 15 stored at 4 • C until measurement. DOC was measured with a Shimadzu Total Organic Carbon Analyzer (TOC-VCPH). An average of three to five injections per sample was used as the measured value. The detection limit for the DOC measurement is 0.25 mg L −1 and the uncertainty of the measurement was ±10 % for values higher than 1.5 mg L −1 , and for values lower than 1.5 mg L −1 the uncertainty was ±15-20 %.

Stable water isotopes 20
To measure stable water isotopes (δ 18 O, δD), 10 mL of the untreated water sample was filled in 10 mL PE narrow-neck bottles. Samples with salinity higher 300 µS cm −1 were measured with an Isotope Ratio Mass Spectrometer (IRMS: Finnigan Delta-S), using equilibration techniques (Meyer et al., 2000). Whereas samples with low salinity were measured with an Ultra High-Precision Isotopic Water Analyzer (PICARRO L2130-i, coupled with an autosampler and vaporizer using Cavity Ringdown Spectroscopy). The internal precision of the H and O isotope measurements is better than ±0.8 ‰ and ±0. 10 ‰, 25 respectively. The oxygen and hydrogen isotopic compositions are given relative to Standard Mean Ocean Water (VSMOW) using the conventional δ-notation.
Stable water isotopes have been widely used in palaeoclimate and palaeohydrological research as isotope fractionation is temperature-dependent. The mean annual δ 18 O of precipitation is positively correlated with the mean annual air temperature, and hence snow typically has a strong isotope variability as well as relatively low (or light) values, particularly at high latitudes 30 (Dansgaard, 1964). Stable water isotopes can be also used to trace water phase changes i.e. during freezing as these are accompanied by kinetic isotope fractionation processes (Souchez and Jouzel, 1984;Lacelle, 2011). An δ 18 O-δD plot gives  (Lacelle, 2011). The extent of fractionation in the system water-ice critically depends on the velocity and rate of freezing (Gibson and Prowse, 1999;Tranter, 2011) which in turn is directly connected to the thermal conditions and the water availability of a given system. Isotopic fractionation during freezing is accompanied by heavier isotope composition 5 for the first ice and lighter isotope composition for the last ice formed Prowse, 1999, 2002). In this study, we differentiate between open and closed system freezing. In an open system, the water source under the ice and hence the isotope composition of the ice formed both remain largely constant, differing from closed-system freezing where the isotopic composition of the water pool changes prior to freezing. Furthermore, the water isotopic signature may be indicative for the mixing of different water masses (endmembers) i.e. precipitation with surface water. The isotopic signature is then indicative 10 for the relative contribution of each endmember and preceding natural isotopic fractionation processes, which changes if the endmembers' specific isotope composition differs. The δ 18 O-δD values in ice may also be directly influenced by precipitation, if liquid or solid precipitation falls on the ice layer and freezes to a part of the ice. In a co-isotope plot, δ 18 O and δD of precipitation generally have values that lie on or near to the Global Meteoric Water Line (GMWL) (Craig, 1961).

Carbon isotopes of methane
The carbon isotopic composition of methane (δ 13 C CH4 ) was measured on the same day and at the same bottle as the methane concentrations, to assure comparability of the data. After measuring the methane concentration, 20 mL of N 2 were added to the sample bottle to increase the headspace of the bottle for stable carbon isotope measurements. The bottle was shaken for at least 30 minutes. 20 mL of gas were removed with a glass syringe by adding 20 mL of Milli-Q water at the same time to 30 equilibrate pressure. δ 13 C CH4 was determined using a Delta XP plus Finnigan mass spectrometer. The extracted gas was purged and trapped with PreCon equipment (Finnigan) to pre-concentrate the sample. The carbon isotopic ratios are given relative to the Vienna Pee Dee Belemnite (VPDB) standard using the conventional δ-notation. The analytical error of the analyses is ±1.5 ‰ for δ 13 C CH4 values. Methane concentration in nanomolar (nM) was calculated with the Bunsen solubility coefficient of Wiesenburg and Guinasso Jr (1979).
For all water bodies, a Rayleigh distillation model of the type discussed by Coleman et al. (1981) and used by Damm et al. (2005Damm et al. ( , 2015 was calculated: where α is the isotope fractionation factor, f is the fraction of the methane remaining and (δ 13 C CH4 ) 0 is the initial isotopic composition. For the Rayleigh model, bacterial oxidation was assumed to be the only methane sink, with no further inputs or mixing that would affect the isotopic composition of methane (Mook, 1994).

Bubble transect
To gain insight into the type and spatial distribution of methane bubbles trapped in the ice of GL a methane bubble transect was 10 mapped. Snow was cleared from a lake ice transect of 70 m length and 1.4 m width starting 30 m from the northwest shoreline aiming towards the lake centre ( Fig. 6). A GoPro Hero 7 camera was used to take densely overlapping photos of the transect from approx. 1.7 m vertical height along the entire transect to photographically record the methane bubble patches captured in the lake ice that are associated with certain seeps types. A measurement tape at the side of the ice-free transect area served as scale in the images. The photos were rectified and mosaicked in the image rectification software AGISoft Professional and 15 the georeferenced mosaic was imported to a desktop geographical information system (ArcGIS, version 10.4). Methane seep types were classified and mapped following Walter Anthony and Anthony (2013). The distance to the shoreline of seep classes was calculated in ArcGIS using the "near" function. The Kernel density estimation in 14 distance classes was calculated in the R environment.

Ice morphology
We compared the data within ice cores as a function of depth below the ice surface, and between the cores of the water bodies.
For a simple comparison between ice cores and the three water bodies, mean values and the range (min., max. values) were calculated for every location (Tab. 2). The length of the ice cores, according to the thickness of the ice cover during field work ranged from 110 to 197 cm. Mean lengths were 144 cm in TB, 166 cm in PF, and 160 cm in GL. The ice of all cores from TB 25 was nontransparent, with a high concentration of enclosed gas bubbles. The uppermost 3 cm (core 27) to 10 cm (core 29) was idenitfied as regelation ice from snow melt. The snow thickness ranged from 11 cm (core 27) to 32 cm (core 29). For PF, the ice of the two cores was clear and contained inclusions of elongated gas bubbles. The uppermost 3 cm of core 32 appeared milky-white. The difference in snow thickness above the two cores was quite large with 8 cm (cores 31) and 20 cm (cores 32).
For GL, the ice morphology was heterogeneous. Core 20 appeared transparent-white, while core 21 included small, elongated

Dissolved organic carbon (DOC)
For TB, the DOC concentration ranges from 1.0 to 2.9 mg L −1 (Tab. 2). The concentrations increase slightly with depth in the upper portion of the cores. At the lower portion, the concentrations decreased, but at the bottom depths they increased 10 again (Fig. 3). For the cores of PF, DOC concentrations increase with depth, with values up to 3.6 mg L −1 in the lower ice. In  (Fig. 4).

Spatial bubble distribution
In the total snow-cleared transect area of 89.2 m 2 we found 29 seeps of class A and five seeps of class B (Fig.6), but none of larger classes C or hot spots (Walter Anthony and Anthony, 2013). The average density of class A seeps in the observed area is 0.33 seeps per square metre. Seep density of class B was more than 6 times lower (0.05 seeps per square metre). Total seep density over the transect (classes A+B) was 0.39 seeps per square metre. The distribution of seep density along the transect 5 shows no linear or homogeneous pattern.

Stable carbon isotopes
In the ice cores from TB, δ 13 C CH4 values ranged from −51.9 ‰ to −36.9 ‰. The values are in a smaller range than the δ 13 C CH4 values for the cores of the other locations (Tab. 2). In PF, the δ 13 C CH4 values range from −79.7 to −31.8 ‰ for both cores (Tab. 2). The cores indicate a similar pattern, with carbon isotopes more enriched in δ 13 C CH4 in the lower portion of the cores  where values ranged from −91.6 to −12.3 ‰, with a strong variability within and between the two cores. Greater variability was observed for methane concentrations.

Discussion
A seasonal ice cover is a barrier to gas exchange between water and the atmosphere. The coastal water bodies studied here are covered by ice for 9 months of the year, a period that is shortening for both lake and marine ice (e. g. Günther et al., 2015). At 10 the same time, air temperatures (Boike et al., 2013), sea level rise rates (Nerem et al., 2018), and coastal thermo-erosion rates (Günther et al., 2013) all increase, indicating a rapidly shifting regime for aquatic environments in the region. The importance of the persistence and duration of the seasonal ice cover in all of these processes is poorly understood. It may act not only as a barrier, but also as a source or sink for methane or as a habitat for microbes that facilitate methane consumption. As further warming of the Arctic shortens the duration of ice cover, pathways to methane emissions will probably shift. The three water bodies in this study represent 1) a terrestrial thermokarst lake (GL), 2) a thermokarst lake that has become a coastal lagoon via thermo-erosion (PF) and the marine shoreface, in a setting where ice-rich Yedoma permafrost is undergoing thermo-erosion (TB). Although the ice cover sampled from all three settings was largely clear ice, important geomorphological differences between sites necessarily should lead to differences in methane dynamics and hydrochemical characteristics. In the lagoon setting, thickening of the ice cover eventually plugged the shallow connection of the brackish basin to the sea, at which point 5 freezing concentrated the remaining brine beneath the ice for the rest of the winter (semi-closed system), a typical situation in lagoons, behind barrier islands or on gently inclined shorefaces. Landfast ice on the shoreface does not close off the water basin, which continues to undergo exchange with the central Laptev Sea and Lena River inflow (open system). The thermokarst lake had neither outlets nor significant inlets, and the ice effectively closed off the water body (closed system), isolating the freshwater basin and its talik. Based on our data, we suggest that the type of water body also determines the circulation 10 of methane. Our results have consequences for the methane source-sink balance in polar terrestrial aquatic systems. In the following we examine each system's ice growth, methane concentrations and compositions and discuss the processes involved.

Tiksi Bay -the open system
Tiksi Bay is part of Buor Khaya Bay and via the central Laptev Sea perennially connected to the Arctic Ocean. The marine impact delays the onset of ice formation compared to the terrestrial aquatic system shown by satellite images from ESA 15 Sentinel-1 and -2. In addition to sea water, snowmelt in spring, small coastal catchments and the Lena River enter into TB between Cape Muostakh and Muostakh Island and around the southern end of Muostakh Island. Lena River discharge follows a nival discharge regime, with very high discharge in the spring and early summer months (Magritsky et al., 2018). While in winter when the connection between Buor Khaya Bay and TB is restricted by sea ice, the contribution of Lena discharge must be much smaller. When TB freezes, it is supposed to be a continuously open system, i.e. water exchange is ongoing during 20 winter below the ice. Both aspects (open system and the mixing of fresh and brackish water) are corroborated in the stable water isotopes composition of the ice cores. A mean δ 18 O value of −15 ‰ for TB is well below full marine conditions and displays the continuous and strong influence of freshwater supply through the Lena River to the Laptev Sea.
Firstly, in an open system such as TB, the water circulates freely beneath the ice cover, impeding the enrichment of lighter water isotopes in the remaining water. Therefore, the isotope composition of the initial ice should remain more or less constant, 25 and hence also that of the ice with depth (Gibson and Prowse, 1999), assuming the freezing velocity is roughly constant.
Accordingly, the water isotopic composition and salinity values are stable until the ice is approximately 80-90 cm thick (Fig.   3,4). The small variation at the top might be due to variability in isotopic fractionation, i.e. related to a change of the freezing rate after ice formation started.
Secondly, TB is influenced by marine and river water, and a change in this ratio may change the isotopic composition of The PF lagoon is on average around 1.5 m deep. At the time of coring the ice thickness was about 1.6 m; consequently the outer lake area, more than 50 % of the lagoon area, was assumed to be frozen to the bed (Strauss et al., 2018). Thermokarst lake basins that are transformed into thermokarst lagoons may be increasingly affected by seawater, at least intermittently, during high water events such as storm surges, resulting in changes to their temperature and salinity regimes (Romanovskii et al., 2000). Increasing salinity in turn also affects subsurface permafrost thaw dynamics and may therefore result in different 5 methane production rates (Angelopoulos et al., 2019). In the uppermost ice, i.e. when freezing started, methane concentration in PF water was more than ten times higher than in TB. The isotopic signature was in the range of microbially produced methane (−70 to −80 ‰) (Whiticar, 1999). This large excess of δ 13 C-depleted methane clearly points to methane from the unfrozen talik, released into the water body and stored therein during the ice-free season.
When PF switched from a connected to a closed system at around 60 cm ice thickness, ongoing methane oxidation beneath 10 the ice lowered the methane concentration captured in the ice to ≤10 nM, comparable to those in TB ice. At the same time, the methane isotopic signature became comparably enriched in δ 13 C (Fig. 5, 9). Ongoing ice formation under closed system conditions (below 60 cm), as indicated by stable water isotopes, induced a continuous increase in ice salinity (Fig. 3) which in turn favoured the shift of the horizon where methane oxidation could occur from the water to the bottom of the ice. This is corroborated by the Rayleigh fractionation curves calculated for ice that grew under closed conditions, using as initial methane 15 isotopic signature the uppermost value, i.e. the signature when freezing began (Fig. 5, 9).
In addition, temperature increases towards the bottom of the ice (Fig. 3). The bottom ice offers a protected environment with favourable conditions for microbial metabolism: relatively warm and stable temperatures, contact with liquid water and permeable ice, permitting migration of gases and nutrients, similar to marine ice, where most bacteria are located in the lowest centimetres of the ice (Krembs and Engel, 2001). During freezing of the ice cover, its growth rate decreases (cf. Anderson,20 1961), providing more time and space for bacterial metabolism. Methane uptake from the water into the bottom of the ice and its oxidation there may have continued over the winter until the ice break-up. Methane oxidation ceases when concentrations are too low for oxidation to be efficient (Cowen et al., 2002;Valentine et al., 2001), at values ranging from 0.6 nM to 10 nM.
Methane concentrations in the ice above 130 cm (Fig. 5) are less than 10 nM, suggesting that ice is an effective sink for methane in this type of water body during winter.

Goltsovoye Lake -the closed system
The ice core hydrochemistry from Goltsovoye reflects a water body that freezes in euqilibrium with atmospheric methane concentrations with two cores showing the influence of snow loading and of ebullition.
Goltsovoye Lake (GL) is an isolated thermokarst basin surrounded by ice-rich Yedoma uplands to the west and east and partially degraded Yedoma uplands to the north and south. The lake is underlain by continuous permafrost some hundreds 30 of meters deep, and has a thaw bulb (talik) beneath its bed due to the positive temperatures at the lake bottom. Water in GL derives from precipitation, most of which falls as snow, overland flow and perhaps as groundwater flow through the shallow active layer. Thus, the concentration of dissolved constituents remains small in lake water and ice, as reflected in very low electrical conductivities of less than 50 µS cm −1 (Fig. 3).
Water depths range from < 1m to about 8.5 m from the western to eastern shore, respectively (Fig. 2). This asymmetric shape influences the progress of ice growth in winter. Ice formation typically starts from the lake shore, most likely at the very shallow west shore, and leads to bedfast ice formation at the position of core 21, i. e. lake ice frozen to the lake bed. Hence, it is likely that core 21 began to form earlier in the season, and that the upper ice in this core, and probably also in core 22, reflects the chemistry of the summer/autumn lake water. All cores except core 23 had similar δ 18 O (around -16.5‰) and δD 5 (-140‰) values, and with a regression slope of 6.6 ( Fig. 8), pointing to freezing under equilibrium conditions (Lacelle, 2011).
At this location both, the methane concentrations and the uniform δ 13 C signature indicate equilibrated values with respect to the atmospheric background values (Fig. 5). This circumstance clearly shows that lake water was not supersaturated during the ice-free season. The other two cores show clear differences and are described in the following.
For core 23 stable isotopes in the upper 120 cm follow the GMWL (Fig. 8), indicating equilibrium fractionation (in the 10 system vapour-water, as snow is involved) due to different precipitation sources. δ 18 O values as low as -28‰ indicate the involvement of snow. The proximity of snow samples to the GMWL is typical for Northern Siberia and has been also found at snow patches on the Bykovsky Peninsula (Meyer et al., 2002). Core 23 was taken proximal to the lake's steepest shoreline, where active thermo-erosion results in shoreline retreat, and that lies in the lee of prevailing winds , where deep snow is expected to accumulate and load the lake ice. 92 cm thick snow lay on the ice at this location at the time of 15 drilling and water streamed out of the hole after coring (Strauss et al., 2018), indicating an ice cover under positive hydrostatic pressure. The ice of core 23 was milky-white from the surface to about 112 to 114 cm depth consistent with a mixture of snow and water. Thus, we conclude a snow signal evident from the stable isotope composition in the upper 120 cm of core 23. The higher EC and DOC in the same interval (relative to lower ice and to cores 20-22 & 24) are unlikely to have derived from snow, however, and imply heterogeneous ice development above 120 cm. These may be the result of mixing of uprising lake 20 water with snow within ice cracks. Adams and Lasenby (1985) describe the formation of white ice (or snow ice) by water percolation through thermally-induced cracks to the surface of the ice, where the water mixes with snow and forms another ice layer above the former ice. This was observed by Adams and Lasenby (1985) when a snow load depressed the surface of the ice cover below the hydrostatic water level. The high and highly variable methane concentrations over this interval, together with the high EC and DOC, suggest that resuspension events, for example slope failure, occurred during ice formation. The 25 carbon stable isotope signature in core 23 may reflect microbial methane oxidation in the sediment rather than oxidation in or beneath the ice.
The highest concentrations of methane were found in core 24 (up to 15 000 nM), with the lowest δ 13 C signatures (down to −150 ‰, Fi. 5). Core 24 lwas drilled above the steepest portion of the lake bed, where ebullition may be a by product of thermokarst processes beneath the bed. The 12 C-enriched signal is consistent with methane that has not been oxidized in the

Conclusions
Methane concentrations in the seasonal ice cover of three types of Arctic water bodies, representing three different stages of permafrost degradation, revealed differences related to the process of ice formation and its importance as mitigator of methane pathways.
In the ice of Tiksi Bay, which is open to the central Laptev Sea throughout the winter and also underlain by permafrost, the 5 stable isotope signatures and electrical conductivity suggest overall brackish conditions, with an increased admixture of riverine waters with ongoing freezing. Ice composition reflected the composition of the upper layer of brackish water throughout the winter and methane concentrations were low but supersaturated.
In the coastal Polar Fox Lagoon, connected to the sea during summer, ice formation and sealing of a connecting channel between lagoon and Tiksi Bay closes off the water body during freezing, isolating and concentrating remaining brackish water 10 beneath the thickening ice during the winter. In the earlier stages of freezing, the lagoon is still connected to Tiksi Bay and shows similar, brackish conditions in stable isotopes and electrical conductivity. In the later stages of freezing, the lagoon is separated from the bay's influence and behaves like a closed system with decreasing δ 18 O and δD and increasing d excess values. Methane is present at variable concentrations in the lagoon, but the concentration profile over depth and the stable isotope signatures suggest that bacterial oxidation takes place at the interface between ice and lagoon water, reducing the methane concentration preserved in the ice.
In a land-locked thermokarst lake surrounded by Yedoma landscapes, rather uniform δ 18 O and δD values and very low electrical conductivity in all lake ice cores (except for one) indicate either subsurface contributions to the lake in winter or a lake deep enough not to behave like a closed system. The exceptional core had a clearly meteoric (likely due to snow-loading 5 of the ice) contribution. Methane concentrations in the lake ice were spatially highly variable. High methane concentrations were local and probably associated with ebullition and snow loading of the ice at an eroding shoreline.
Overall, ice on coastal waters in this environment acts primarily as a barrier to methane fluxes to the atmosphere, a barrier that is effective for most of the year but also will be effected by rapid changes due to Arctic warming and associated ice thinning. Additionally, we have shown that the ice cover may act as a sink, providing a habitat for methane oxidation.