Heterogenous CO2 and CH4 content of glacial meltwater of the Greenland Ice Sheet and implications for subglacial carbon processes

Accelerated melting of the Greenland Ice Sheet (GrIS) has increased freshwater delivery to the Arctic Ocean and amplified the need to understand the impact of GrIS meltwater on Arctic greenhouse gas (GHG) budgets. We measured carbon 10 dioxide (CO2) and methane (CH4) concentrations and d13C values and use geochemical models to evaluate subglacial CH4 and CO2 sources and sinks in water discharging from three subglacial outlets of the GrIS in southwest (Isunnguata and Russell Glaciers) and southern Greenland (Kiattut Sermiat). CH4 concentrations vary by orders of magnitude between sites and are saturated with respect to atmospheric concentrations at Kiattut Sermiat, but are supersaturated at southwest sites, even though oxidation reduces concentrations by up to 50% during periods of low discharge. CO2 concentrations range from supersaturated 15 at Isunnguata to undersaturated at Kiattut Sermiat. CO2 is consumed by mineral weathering throughout the melt season at all sites, however differences in the magnitude of subglacial CO2 sources result in meltwaters that are either sources or sinks of atmospheric CO2. The predominant source of CO2 at Isunnguata is organic matter (OM) remineralization, but Russell and Kiattut Sermiat sites have multiple or heterogeneous subglacial CO2 sources that maintain atmospheric CO2 concentrations at Russell but not at Kiattut Sermiat where CO2 is undersaturated. These results highlight the variability in GHG dynamics under 20 the GrIS. Constraining this variability will improve our understanding of the impact of GrIS melt on Arctic GHG budgets, as well as the role of continental ice sheets in GHG variations over glacial-interglacial timescales.


Introduction
Glaciers play an important role in global chemical cycles due to the production of fine-grained sediments that participate in carbonate and silicate mineral weathering reactions (Table 1), which are the principal sink of atmospheric CO2 25 over geologic timescales (Berner et al., 1983;Walker et al., 1981). Variations in the weathering intensity of comminuted sediments may contribute to glacial-interglacial atmospheric CO2 variations as sediments are alternately covered by ice and exposed following ice retreat. However, the importance of CO2 consumption by mineral weathering is poorly understood, including effects from the advance and retreat of continental ice sheets (Ludwig et al., 1999). Recent evaluations of carbon budgets in proglacial environments indicate that mineral weathering results in net sequestration of atmospheric CO2, suggesting that proglacial systems are underrecognized as Arctic CO2 sinks (St. Pierre et al., 2019), however alternate processes could lead to the production of greenhouse gases (GHG) in glacial systems. For instance, CH4 production in anaerobic subglacial environments driven by the remineralization of organic matter (OM) contained in soils and forests covered during glacial margin fluctuations has been suggested as a potential carbon feedback to drive warming (Sharp et al., 1999;Wadham et al., 2008). Because the global warming potential of CH4 is 25 times greater than CO2, even limited subglacial methanogenesis has 35 the potential to strongly impact the GHG effect of glacial meltwater. Combined inorganic and organic subglacial processes may therefore produce glacial meltwater that is a source or sink of GHG. While the net impact of these processes on modern carbon fluxes is poorly constrained, determining these impacts will improve modern carbon budgets as well as depictions of how fluxes may have evolved during the advance and retreat of continental ice sheets.
In subglacial environments where remineralization is limited by low OM availability, the major element solute load 40 of glacial meltwater is typically dominated by products of mineral weathering reactions (Tranter, 2005). The extent of mineral weathering in subglacial environments depends in part on the availability of acids to drive reactions, namely sulfuric and carbonic acids (Table 1). Sulfuric acid is derived from the oxidation of reduced sulfur species, which largely occur as ironsulfide minerals including pyrite (Tranter, 2005). Sulfide oxidation may occur abiotically, however the kinetics of microbially mediated sulfide oxidation is several orders of magnitude faster and may lead to local depletion of oxygen provided sufficient 45 supply of sulfide minerals (Sharp et al., 1999). In contrast, carbonic acid may be derived from multiple external or in situ sources of CO2 to the system. The dominant external source is supraglacial meltwater that flows to the subglacial system through moulins following equilibration with atmospheric CO2 (Fig. 1). Unlike proglacial environments where free exchange between water and the atmosphere may resupply CO2 consumed by weathering, subglacial environments may be partially or fully isolated from the atmosphere, limiting further atmospheric CO2 invasion and thus the extent of mineral weathering with 50 carbonic acid. However, additional atmospheric CO2 may be delivered in open portions of the subglacial environment though exchange in fractures or moulins along subglacial flow paths or in partially air-filled conduits, allowing a much greater magnitude of carbonic acid mineral weathering (Graly et al., 2017). CO2 may also be derived from in situ sources, such as gaseous CO2 contained in ice bubbles of basal ice, or fluid inclusions in rocks that release volatiles (including CO2) following mechanical grinding (Macdonald et al., 2018). When OM is available, its remineralization also generates CO2 (and potentially 55 CH4) along with nutrients, but low OM availability in many subglacial systems limits remineralization as a CO2 source (Fig.   1).
The role of subglacial carbon processes may play an increasingly important role in modern Arctic carbon budgets as disproportionate warming increases glacial meltwater and sediment fluxes to the ocean, particularly from the Greenland Ice Sheet (GrIS). The Greenland Ice Sheet (GrIS) is the last remaining ice sheet in the Northern hemisphere following collapse of 60 all others since the Last Glacial Maximum (~25 ka). It has been losing mass at increasing rates that averaged 286±20 Gt/yr between 2010-2018, representing a six-fold increase since the 1980s (Mouginot et al., 2019). While mineral weathering significantly modifies the composition of GrIS subglacial discharge (e.g. Hindshaw et al., 2014;Deuerling et al., 2018;Urra https://doi.org/10.5194/tc-2020-155 Preprint. Discussion started: 3 July 2020 c Author(s) 2020. CC BY 4.0 License. et al., 2019) and should consume CO2 similar to other glacial and proglacial environments, the recent identification of microbially driven reactions (including methanogenesis) in subglacial environments of the GrIS indicates that organic 65 processes may also play a role (Dieser et al., 2014;Lamarche-Gagnon et al., 2019;Musilova et al., 2017). The relative importance of subglacial GHG sinks (CO2 consumption through mineral weathering) and sources (such as OM remineralization) determine the GHG composition of subglacial discharge, which may then serve as a source or a sink of atmospheric GHGs. Constraining the relative impacts and variability of these processes underneath the GrIS will provide important information regarding the current and future impact of GrIS loss on Arctic carbon budgets, as well the role of 70 continental ice sheets on carbon cycle feedbacks.
To evaluate the net impact of carbon processes on the GHG composition of subglacial discharge of the GrIS, we compare water chemistry, dissolved CO2 and CH4 concentrations, and gas stable isotopic compositions between three subglacial discharge sites draining land-terminating glaciers of the GrIS over the melt seasons of 2017 and 2018 (Fig. 2). We employ mass balance models utilizing the concentrations of major cations and anions to determine the magnitude of the impact 75 on CO2 concentrations from mineral weathering reactions (Table 1). These results are combined with measured gas concentrations to determine the relative importance of mineral weathering compared to OM remineralization on the CH4 and CO2 content of subglacial discharge. We also assess the temporal and spatial variability of these processes under the GrIS to improve our understanding of carbon cycling in Greenland subglacial environments and the implications of GrIS mass loss on Arctic carbon budgets. 80

Study locations
Our three subglacial discharge locations are located in southwest (Fig. 2a,b) and southern (Fig. 2a,c) Greenland. The Isunnguata Glacier (IS; 67°09'27.1" N, 50°03'25.0" W) and Russell Glacier (RU; 67°05'22.1" N, 50°14'18.7" W) drain to the Watson River, which is one of the largest proglacial rivers in Greenland. Watson River discharge is monitored by PROMICE 85 (Programme for Monitoring of the Greenland Ice Sheet; van As et al., 2018) and total discharge was 4.3 and 3.6 km 3 of water in 2017 and 2018, respectively . The total catchment size for the Isunnguata is 15,900 km 2 , though our water samples were collected from a smaller sub-catchment with a drainage area of approximately 40 km 2 (Lindbäck et al., 2015;Rennermalm et al., 2013). The total drainage area for the Russell glacier is estimated at 300 km 2 . This estimate comes from subtracting the Leverett drainage area estimated at approximately 600 km 2 (Hawkings et al., 2016) from the 90 approximately 900 km 2 catchment that includes the Russell and Leverett drainages (Lindback et al., 2015). While discharge from the third site in southern Greenland, Kiattut Sermiat (KS; 61°12'13.5" N, 45°19'49.1"W), is not monitored, a previous study using dye tracing techniques estimated approximately 0.22 km 3 of discharge in 2013, and its catchment size was estimated at 36 km 2 (Hawkings et al., 2016). Underlying lithologies differ between sites. Watson River sites are located near the boundary between the Archean Craton to the south and the southern Nagssugtoqidian Orogen to the north (Henriksen et al., 2009). The Archean block is composed of granites and granulite facies orthogneisses that were intruded by mafic dykes during Paleoproterozoic rifting. These rocks were deformed and modified during subsequent continent-to-continent collision in the Paleoproterozoic to create the amphibolite facies gneisses of the southern Nagssugtoqidian Orogen (van Gool et al., 2002). Kiattut Sermiat lies within the Paleoproterozoic Ketilidian fold belt (Henriksen et al., 2009). Lithologies in this region include the Julianehåb Granite and associated basic intrusions and the sedimentary and volcanic rocks of the Mesoproterozic 100 Gardar Province that include a suite of alkaline igneous rocks and basaltic dykes with interbedded sandstones (Kalsbeek and Taylor, 1985;Upton et al., 2003).
Previous studies have characterized chemical weathering reactions in subglacial discharge to the Watson River (Deuerling et al., 2018;Yde et al., 2014), Kiattut Sermiat (Hawkings et al., 2016), and comparatively between these sites (Urra et al., 2019). There has been extensive work regarding ice sheet dynamics and hydrology in the 105 Watson River catchment (Van As et al., 2017Lindbäck et al., 2015) as well as Kiattut Sermiat (Warren and Glasser, 1992;Winsor et al., 2014). Previous studies near these study locations have identified CH4 and CO2 supersaturation in subglacial discharge of the Isunnguata site (Christiansen and Jørgensen, 2018;Ryu and Jacobson, 2012), while methanogenic microbial communities have been observed at Russell Glacier (Dieser et al., 2014) and CH4 supersaturation at the Leverett Glacier (Lamarche-Gagnon et al., 2019), which also flows into the Watson River (Fig. 2c). Subglacial permafrost has been 110 identified near the Isunnguata site (Ruskeeniemi et al., 2018) and attributed to Holocene fluctuations in the ice sheet margin.
While a similar Holocene ice retreat and re-advance may have occurred in southern Greenland (Larsen et al., 2016), it is unknown whether this retreat led to the formation of organic deposits.

Sample collection
We collected water samples from subglacial discharge sites in spring and fall of 2017, and the summer of 2018 to 115 observe seasonal variations in water composition. Samples were collected as close as possible to the subglacial discharge site, which was less than 10 m for the Isunnguata site, approximately 100 m for the Russell glacier site, and approximately 1.1 km for the Kiattut Sermiat site (Fig. 2). We collected water samples by pumping water through a 0.5-cm flexible PVC tube that was placed in flowing water as far as possible from shore (approximately 1-2 m). A YSI Pro-Plus sensor that was calibrated daily was installed in an overflow cup filled from the bottom to measure specific conductivity (Sp.C), temperature, pH, 120 dissolved oxygen, and oxidation-reduction potential (ORP). These parameters were monitored until stable, between about 10 and 30 minutes, after which samples were collected and preserved in the field according to the solute to be measured after being filtered through a 0.45 µm trace-metal grade Geotech high capacity disposable canister filter. Samples for cations and anions were collected in HDPE bottles; cation samples were preserved with Optima-grade ultrapure nitric acid (pH<2) while no preservative was added to anion samples. Samples for ammonium (NH4) were filtered into 15 mL polypropylene containers 125 and frozen until analysis. Dissolved inorganic carbon (DIC) samples were filtered through 0.2 µm filters directly to the bottom of 20 ml Qorpac glass vials and allowed to overflow until sealed tightly with no headspace. https://doi.org/10.5194/tc-2020-155 Preprint. Discussion started: 3 July 2020 c Author(s) 2020. CC BY 4.0 License.
Gas samples were collected in duplicate via headspace extractions according to methods outlined in Repo et al. (2007) and Pain et al. (2019). Unfiltered water was pumped into the bottom of 500 mL bottles until they overflowed. Bottles were immediately capped with rubber stoppers fitted with two 3-way inlet valves. 60 mL of water was extracted from one inlet and 130 replaced with 60 mL of atmospheric air (for spring and fall 2017 sampling trips) or ultrapure N2 gas in a gas bag (summer 2018 sampling trip). Bottles were shaken for 2 minutes to equilibrate headspace gas with water, and headspace gas was extracted and immediately injected into 60 ml glass serum bottles that had been evacuated immediately prior to sample introduction. Samples were stored at room temperature until analysis, which occurred within one week of collection. Measured headspace concentrations were converted to dissolved concentrations using methods outlined in Pain et al. (2019). When 135 atmospheric air was used for headspace extractions, atmosphere samples were collected in tandem and analyzed to correct each sample for calculated dissolved CO2 and CH4 concentrations and isotopic compositions. This correction altered CH4 concentrations by up to 22% for one sample from the Russell glacier, though less than 5% for all other samples, and resulted in a correction of d 13 C-CH4 of up to 1.3‰. For CO2, the correction altered concentrations by up to 15% for one sample collected at Kiattut Sermiat, though less than 10% for all other samples, and resulted in a correction of d 13 C-CO2 of up to 0.4‰. 140 For fall 2017 and summer 2018 sampling trips, alkalinity was measured in the field laboratory within 3 days of collection by titration with 0.01 N HCl using the Gran method. Because alkalinity measurements were not available for the spring 2017 sampling trip, we estimate alkalinity with PHREEQc modeling and the phreeqc.dat database (Parkhurst, 1997) using major cations and anions, pH, temperature, and DIC concentrations as model inputs.

Laboratory analysis 145
Gas samples were analyzed for CO2 and CH4 concentrations, and d 13 C-CO2 and d 13 C-CH4 on a Picarro G2201-i cavity ring-down spectrometer. Carbon isotopic compositions are reported in reference to Vienna Pee Dee Belemnite (VPDB). Check standards of known CO2 and CH4 concentrations and isotopic compositions were measured during each sample run and were accurate within 10%. Anion and cation concentrations were measured on an automated Dionex ICS-2100 and ICS-1600 Ion Chromatograph, respectively. Error on replicate analyses was less than 5%. DIC concentrations were measured on a UIC 150 (Coulometrics) 5011 CO2 coulometer coupled with an AutoMate Preparation Device. Samples were acidified and the evolved CO2 was carried through a silver nitrate scrubber to the coulometer where total C was measured. Accuracy was calculated to be ±0.1 mg/L based on measurement of check standards. Dissolved ammonium (NH4) concentrations were analyzed on a Seal AutoAnalyzer III. Error on check standards was less than 10%.
Values of ec reflect methanogenesis pathways (acetoclastic or CO2 reduction) as well as the extent of oxidation.
Values of ec between approximately 40 and 55‰ are produced for CH4 produced via acetoclastic methanogenesis, while CO2 160 reduction produces values between approximately 55 and 90‰. Lower values (ec between 5 and 30) result when CH4 oxidation predominates. Modern atmospheric input without additional alteration of CO2 or CH4 isotopic systematics results in a ec value of approximately 40 (Whiticar, 1999).
We calculated CH4 oxidation using the isotopic method outlined in Mahieu et al. (2008) and Preuss et al. (2013). The fraction of oxidized methane (fox) in an open system is given by: 165 where dE is the measured d 13 C-CH4 value for each water sample, dP is d 13 C-CH4 of produced methane, αox is the oxidation fractionation factor, and αtrans is a fractionation factor resulting from diffusive transportation of CH4. While the exact value of dP is unknown, diagenetic alteration of d 13 C-CH4 values through oxidation or transport only enrich d 13 C-CH4 signatures, therefore the value of dP is taken as the most depleted d 13 C-CH4 signature assuming it is the least impacted by diagenetic alteration. Literature-reported values for αox range between 1.003 and 1.049. We calculate the fraction of oxidized methane 170 with the largest fraction factor (αox = 1.049; Mahieu et al., 2008), which yields the minimum amount of CH4 oxidation required to explain the observed variations in d 13 CH4, and thus is a conservative estimate for CH4 oxidation. Literature-reported values for αtrans vary from 1 for advection-dominated systems to 1.0178 for diffusion-dominated porous media (Visscher et al., 2004;Mahieu et al., 2008;Preuss et al., 2013). We assume that transport is advection dominated and thus assume αtrans = 1.

Mineral weathering and carbonate modeling 175
We used major cation and anion concentrations and alkalinity to partition solutes into the four mineral weathering reactions in Table 1 after correcting solute concentrations for marine aerosol deposition using measured chloride concentrations and standard seawater element ratios. The mass balance model followed the methods of Deuerling et al. (2019).
After apportioning solutes to mineral weathering reactions, we used the stoichiometries of reactions to calculate the impact of each reaction on dissolved CO2 concentrations (Table 1). The mineral weathering model apportions solutes to reactions in 180 Table 1 based on the ratios of Ca/Na and Mg/Na in silicate minerals in stream bedload samples, which were taken to be 0.54 and 0.38, respectively, for Isunnguata and Russell Glacier samples (Deuerling et al., 2019;Hindshaw et al., 2014;Wimpenny et al., 2010Wimpenny et al., , 2011 as well as the total impact of mineral weathering on CO2 concentrations (Total CO2-MW), https://doi.org/10.5194/tc-2020-155 Preprint. Discussion started: 3 July 2020 c Author(s) 2020. CC BY 4.0 License.
where changes in the concentrations of CO2 are defined by their absolute values. To discuss the relative importance of individual reactions, we define proportional contributions of each reaction as follows: We combine measured CO2 concentrations with Net CO2-MW in order to determine the magnitude of CO2 production or consumption in the subglacial environment due to processes besides mineral weathering. This analysis assumes that the 195 concentration of CO2 measured at the subglacial outlet is equivalent to the net change in CO2 due to mineral weathering plus the sum of all other subglacial CO2 sources and sinks. We refer to the sum of all other subglacial CO2 sources and sinks as CO2-total, which represents the amount of CO2 that must have been supplied to the subglacial environment to balance the mineral weathering CO2 sink: The sources of CO2 to CO2-total may be evaluated through the use of Keeling plots, which are constructed as the inverse of CO2 concentrations ([CO2] -1 ) versus stable isotopic composition (d 13 C-CO2). If variations in the concentration and isotopic composition of CO2 arise from the mixing of two CO2 reservoirs with constant isotopic compositions and concentrations (Keeling, 1958), a linear relationship is expected between [CO2] -1 and d 13 C-CO2. The y-intercept of a regression between these variables represents the isotopic composition of the high-CO2 end member. Because measured CO2 concentrations include 205 both subglacial CO2 sources and sinks, which may include considerable consumption through mineral weathering reactions, the magnitude of the total subglacial CO2 source is taken as CO2-total. We therefore construct Keeling plots between [CO2-total] -1 and measured d 13 C-CO2 values because while mineral weathering impacts the concentration of CO2, its isotopic composition is not appreciably altered (Myrttinen et al., 2012) compared to the range of isotopic compositions of potential CO2 end members, namely OM remineralization, atmospheric CO2, and lithogenic CO2 sources due to mechanical grinding (Fig. 1). 210

Discharge relationships
We evaluate the relationship between subglacial CH4 and CO2 dynamics and discharge using discharge records provided by PROMICE (Programme for Monitoring of the Greenland Ice Sheet; van . Discharge records are collected at the outlet of the Watson River, which represents the combined discharge of Isunnguata and Russell glaciers as well as other outlet glaciers including the Leverett Glacier and major tributaries including the Orkendalen River (Fig. 2).
Watson River discharge estimates are therefore greater than the true discharge of our individual sampling locations, however we assume that discharge at the Isunnguata and Russell subglacial outlet sites is roughly proportional to Watson River discharge and exhibits similar temporal variability (Rennermalm et al., 2012). Because diurnal fluctuations in river discharge can be large, and differing water travel times from subglacial outlet sites to the Watson River mouth induces a lag of up to 8 hours between maximum daily discharge at subglacial discharge sites and the Watson River outlet, we compare subglacial 220 CH4 and CO2 concentrations to average daily discharge, calculated as the average of hourly discharge estimates over the days on which subglacial discharge water samples were collected. Because no discharge information is available for Kiattut Sermiat, we assess discharge relationships at Watson River (Isunnguata and Russell) sites only.

Temporal variability in water chemistry and gas concentrations 225
Chemical parameters differ between subglacial discharge sites as well as throughout the melt season. Specific conductivity (Sp.C; Fig. 3a) is typically highest at Kiattut Sermiat (26±8 µS/cm), followed by Russell (22±5 µS/cm) and and Kiattut Sermiat (-16.1±1.6‰) sites, though similar seasonal variation occurs for all sites with relatively more depleted values in the spring and fall compared to summer.

Methane oxidation and relationship with discharge 245
Values of ec are similar throughout the melt season for Isunnguata (38±10‰) and Russell (38±9‰) and are relatively higher in the summer sampling period, while Kiattut Sermiat ec values are higher on average (42±13‰; Fig. 4a) with lowest values in the summer. Estimates of fox are similar between Isunnguata (17±15%), Russell (23±15%), and Kiattut Sermiat sites (25±22%; Fig. 4b). However, fox values are higher in the spring and fall sampling times compared to summer for Isunnguata and Russell and approach 50% in the spring, while Kiattut Sermiat values decrease throughout the melt season. 250 CH4 concentrations are unrelated to Watson River average daily discharge for Isunnguata, but significantly negatively correlated for Russell (Fig. 5a). The fraction of CH4 oxidized is moderately negatively correlated with discharge for both Isunnguata and Russell, though the relationship is not significant for either site (Fig. 5b). Discharge is positively correlated with ec for both Isunnguata and Russell (Fig. 5c). While the relationship is only significant for Isunnguata, the slopes and intercepts of regression lines for Isunnguata and Russell are similar. 255
Because Isunnguata samples exhibit a linear relationship between [CO2-total] -1 and d 13 C-CO2 consistent with two endmember mixing, we utilize the d 13 C-CO2 of samples and end members defined by the Keeling plot in an isotopic mixing model to calculate the relative contributions of CO2 sources to CO2-total. We take the y-intercept (-27.5‰; Fig. 8a) of the Keeling plot regression as the d 13 C-CO2 for the high-CO2 end member (assumed to be generated by OM remineralization; CO2-OM) and assume atmospheric CO2 as the low-CO2 end member (CO2-atm; d 13 C-CO2 = -8‰). The relative (%CO2-atm and %CO2-280 OM) and absolute (CO2-atm and CO2-OM) contributions of these CO2 sources to Isunnguata CO2-total vary throughout the melt season, with CO2-OM being the dominant source in the early and late melt season while CO2-atm is the dominant source in the mid-melt season (Fig. 9a). The absolute magnitude of CO2-OM is approximately 5-10 times greater in the early (89 µM) and late (117 µM) melt seasons compared to mid-melt seasons (13-23 µM), while the magnitude of CO2-atm varies relatively little throughout the melt season at around 50 µM (Fig. 9b). Both %CO2-atm and %CO2-OM are significantly (p<0.05) correlated with 285 Watson River discharge, and the correlation is positive for %CO2-atm and negative for %CO2-OM (Fig. 9c). The magnitude of CO2-OM exhibits a power-law relationship with discharge and is highest at lowest discharge, while the magnitude of CO2-atm is invariable as a function of discharge (Fig. 9d).

Discussion
We observe orders of magnitude variability in dissolved CH4 and CO2 concentrations in subglacial discharge of the 290 GrIS, indicating significant differences in the magnitudes of the sources and sinks of these gases across time and space.
Supersaturation of both CO2 and CH4 with respect to atmospheric concentrations indicates that Isunnguata discharge is a source of both gases to the atmosphere, neighboring Russell Glacier discharges water that is a source of CH4 but near equilibrium with respect to CO2, while Kiattut Sermiat in southern Greenland is a sink of atmospheric CO2 but near equilibrium with respect to CH4 (Fig. 3e, g). Because CH4 dynamics may be largely microbially driven while CO2 dynamics include microbial 295 as well as abiotic mineral weathering processes, we first discuss CH4 dynamics including a comparison of concentrations, isotopic compositions, and extent of oxidation between sites and over the melt season. We then discuss CO2 concentrations, impacts of mineral weathering reactions (Table 1), and an assessment of subglacial CO2 sources, including OM remineralization. These assessments will contribute to our understanding of the variability and controls of CH4 and CO2 concentrations in subglacial discharge from the GrIS and may improve predictions of the impact of future ice melt on Arctic 300 carbon budgets.

Sources and sinks of CH4
Differences in CH4 concentrations and relationships with discharge between sites imply heterogeneity in both the terrestrial vegetation (Ruskeeniemi et al., 2018). Methanogenesis fueled by organic material overridden during ice sheet growth has been suggested as a potential climate feedback over glacial interglacial timescales (Wadham et al., 2008), and may contribute to variations in CH4 concentrations between southwest and southern sites in this study.
Subglacial methanogenesis may additionally be controlled by hydrologic factors as the subglacial hydrological network develops throughout the melt season and channelization of meltwater conduits increases subglacial drainage efficiency 320 (Andrews et al., 2015;Cowton et al., 2013). Drainage efficiency impacts both subglacial water residence time as well the transport of aerobic supraglacial meltwater to the ice bed. Both residence time and oxygen delivery may impact subglacial redox status and methanogenesis potential, and favor methanogenesis when oxygen supply rates are low compared to OM remineralization rates. This condition is most likely to be met in distributed subglacial systems that are hydrologically isolated with limited inputs from aerobic supraglacial meltwater. Such a hydrologic control on methanogenesis at the Russell Glacier 325 is supported by the significant negative correlation between CH4 concentrations and Watson River average daily discharge ( Fig. 5a), suggesting that CH4 production occurs predominantly during periods of low discharge and greater residence time.
Alternatively, higher CH4 concentrations during low discharge could result from the dilution of relatively small volumes of methanogenic subglacial meltwater with increasing volumes of aerobic supraglacial meltwater. The lack of relationship between CH4 concentrations and discharge at the Isunnguata Glacier may indicate a greater influence of subglacial outburst 330 events, in which hydrologically isolated methanogenic meltwater pockets are stochastically drained as the subglacial drainage network extends throughout the melt season (Fig. 5a). Subglacial outburst events were also implicated by heterogeneous CH4 concentrations in subglacial discharge of the Leverett Glacier, and could contribute to heterogeneity in CH4 concentrations at the Isunnguata Glacier (Lamarche-Gagnon et al., 2019).
While our results suggest heterogeneity in the extent and controls of methanogenesis between outlet glaciers, the 335 microbial methanogenesis pathway as well as CH4 oxidation dynamics are consistent between sites. Methanogenesis pathways may be evaluated by d 13 C-CH4 as well as ec values because methanogenesis pathways impart different isotopic signatures to CH4 and CO2 (Whiticar and Schoell, 1986). Dieser et al. (2014) measured a microbial d 13 C-CH4 production signal at the Russell Glacier with values between -63‰ and -64‰, which was interpreted to reflect a possible combination of CH4 produced through https://doi.org/10.5194/tc-2020-155 Preprint. Discussion started: 3 July 2020 c Author(s) 2020. CC BY 4.0 License. both acetoclastic and CO2 reduction pathways. The most depleted d 13 C-CH4 value measured at the Isunnguata in this study is 340 close to that of Dieser et al. (2014) at -62.7‰ (Fig. 3f), and similar to values reported by Lamarche-Gagnon et al. (2019) for the Leverett Glacier, suggesting similar methanogenesis pathways across this region. While the exact contributions from each methanogenesis pathway cannot be inferred from isotopic information alone, the range of ec values at outlet glaciers are consistent with predominantly acetoclastic methanogenesis during the peak melt season (Fig. 4a). However, ec values fall below the expected range from acetoclastic methanogenesis during the early and late melt seasons, likely resulting from 345 variations in the extent of subglacial CH4 oxidation. Seasonal variation in CH4 oxidation is supported by consistency between ec and fox values, which both indicate the greatest impact of oxidation (approaching 50%) in the early melt season compared to peak melt season (Fig. 4a, b), with additional evidence of elevated CH4 oxidation in the late melt season at both Isunnguata and Russell glaciers.
The extent of CH4 oxidation may be controlled by multiple factors including oxygen availability, subglacial 350 residences time, and subglacial hydrology, similar to methanogenesis. A hydrologic control of CH4 oxidation is supported by relationships between fox and ec with Watson River daily discharge at both Isunnguata and Russell Glaciers: fox is negatively related with discharge for both sites (Fig. 5b) while ec is positively correlated with discharge (Fig. 5c) although the correlation is only significant at Isunnguata. These correlations suggest that CH4 oxidation is greatest during periods of low flow, which may be associated with greater residence times to allow subglacial CH4 oxidation. The delivery of oxygen to the subsurface 355 by supraglacial melting does not appear to be a limiting factor in subglacial CH4 oxidation, which should increase fox as more oxygenated supraglacial water is delivered to the subglacial system. Instead, the observed greater CH4 oxidation during periods of low discharge may reflect CH4 oxidation following mixing between draining methanogenic subglacial meltwater pockets with aerobic subglacial meltwater. Longer transit times during periods of low flow may allow more subglacial methane oxidation to occur than during peak discharge, when the development of channelized flow paths reduces meltwater residence 360 time in the subglacial environment.
Our results indicate a high degree of heterogeneity in subglacial methanogenesis under the GrIS, as well as a significant impact of CH4 oxidation, which serves to reduce atmospheric CH4 fluxes. Given the observed heterogeneity in this study, further investigation of the spatial variability in outlet glacier CH4 concentrations is needed to determine the impact of GrIS loss on Arctic and global CH4 budgets, while a better understanding of the controls of these differences will improve 365 models of how CH4 fluxes from subglacial discharge will change with continued warming.

Sources and sinks of CO2
Dissolved CO2 concentrations in subglacial discharge are consistently supersaturated with respect to atmospheric concentrations at Isunnguata Glacier, near atmospheric equilibrium at Russell Glacier, and undersaturated at Kiattut Sermiat Glacier, indicating that glacial meltwater from the GrIS can serve as either a source or sink of CO2 to the atmosphere. approach to assess the magnitude of subglacial CO2 sources (including subglacial OM remineralization) depends in part on modelling results of CO2 consumption by mineral weathering (Eq. 10), we first discuss the impacts of mineral weathering reactions, followed by a discussion of subglacial CO2 sources, including OM remineralization.

Subglacial CO2 sink: mineral weathering reactions 375
Although mineral weathering reactions may either increase or decrease dissolved CO2 concentrations (Table 1), the net impact of mineral weathering at our study sites is to consume CO2 (Fig. 6a). Net consumption occurs because the CO2 source from CarbSA is ubiquitously low compared to sinks from either CarbCA or SilCA (Fig. 6b). The range in Net CO2-MW is similar between subglacial discharge sites (between 10-150 µM; Fig. 6a), but average values increase from Kiattut Sermiat to Russell to Insunnguata, likely reflecting the relative weatherability of alkaline igneous rocks, granulite facies gneisses, and 380 amphibolite facies gneisses. Kiattut Sermiat is characterized by a relatively high proportion of CarbCA compared to Watson River sites, which may arise from the presence of trace carbonates in abundant readily weatherable basaltic intrusions as has been implicated in other studies (Urra et al., 2019). The relatively greater influence of carbonate dissolution compared to silicate dissolution on Total CO2-MW at Kiattut Sermiat may also relate to more rapid dissolution kinetics of carbonates, which allow carbonate dissolution to have a large influence on major cation and anion load even when carbonates are only present in 385 trace amounts (Tranter, 2005). At Isunnguata and Russell glaciers, SilCA has a greater influence than CarbCA on Total CO2-MW, which could result from either a lower abundance of trace carbonates to participate in weathering reactions, or relatively longer subglacial residence times that would allow a greater accumulation of silicate weathering products.
Despite the high impact of CarbCA on Total CO2-MW at Kiattut Sermiat compared to Isunnguata and Russell sites, CarbSA is notably lower at Kiattut Sermiat than other sites and suggests a limited role for sulfuric acid weathering that may 390 relate to subglacial sulfide oxidation dynamics. Lower abundances of sulfide minerals in the subglacial environment may limit the production of sulfuric acid, and could result from differences in lithology between sites, the depletion of sulfide minerals due to prior weathering (Graly et al., 2014), or weathering occurring in anoxic environments that limit the oxidation of sulfide to sulfuric acid (Deuerling et al., 2019). The kinetics of sulfide oxidation may also significantly differ between sites depending on the relative contributions of abiotic compared to microbially mediated sulfide oxidation, as microbially mediated sulfide 395 oxidation is several orders of magnitude faster than abiotic sulfide oxidation. Rapid microbially mediated sulfide oxidation has been implicated in the development of anaerobic conditions, which could also support subglacial methanogenesis (Sharp et al., 1999). Observations of higher CH4 concentrations as well as higher contributions of CarbSA at Isunnguata and Russell compared to Kiattut Sermiat in this study may therefore be linked to subglacial microbial activity, which is known to vary based on factors such as the presence of organic and fine-grained rock flour to serve as growth substrates, insulation from 400 fluctuations in temperature, and delivery of nutrients and organic matter from supraglacial sources (Sharp et al., 1999). If microbially driven, our results suggest possible linkages between microbial processes and subglacial mineral weathering regimes, with significant impacts to both CH4 and CO2 dynamics due to the role of CarbSA as a CO2 source (Table 1).

Subglacial CO2 sources
Mineral weathering leads to net CO2 consumption in all subglacial discharge samples, and thus the measured CO2 405 concentrations represents only a fraction of the total CO2 that would have been present in the absence of mineral weathering reactions (CO2-total; Eq. 10). CO2 sources could include dissolution of atmospheric gases in air-filled conduits or fractures in ice, or CO2 contained in ice bubbles ( Fig. 1; Anklin et al., 1995;Graly et al., 2017). CO2 may also be produced through mechanical grinding and volatilization of fluid inclusions (Macdonald et al., 2018) or OM remineralization. While previous studies have indicated that additional atmospheric CO2 input through fractures and air-filled conduits may supply sufficient 410 CO2 to drive the observed extent of mineral weathering in many subglacial environments, including several sites in Greenland (Graly et al., 2017), CH4 concentrations elevated above atmospheric equilibrium at the two Watson River sites reflects OM remineralization that would also contribute CO2. While the magnitude of this source and its relative importance compared to other subglacial CO2 sources is currently unknown, differing sources of carbonic acid for mineral weathering reactions carry different implications for subglacial CO2 budgets. For instance, carbonic acid weathering driven by invasion of atmospheric 415 CO2 would represent a sink of atmospheric CO2, but carbonic acid weathering driven by OM remineralization would instead serve to consume CO2 from in situ sources and limit its potential as an atmospheric source. Determining the sources of carbonic acid to subglacial weathering reactions is therefore critical to understand the controls of mineral weathering in subglacial environments as well as the role of that process in atmospheric CO2 sequestration.
Comparisons between measured d 13 C-CO2 in subglacial discharge samples and likely d 13 C-CO2 values of CO2 sources 420 indicate that CO2 sources differ between sites, with OM remineralization as the predominant CO2 source at the Isunnguata but not at Russell or Kiattut Sermiat glaciers. Keeling plots of [CO2-total] -1 versus d 13 C-CO2 indicate that CO2-total at Isunnguata discharge may be represented by a two-end member mixing model, in contrast to Russell and Kiattut Sermiat glaciers (Fig. 8).
Mixing model end members include a 13 C-enriched, lower-CO2 source and a 13 C-depleted, higher-CO2 source (Fig. 8a). The y-intercept of the regression between [CO2-total] -1 versus d 13 C-CO2 (representing the isotopic signature of the high-CO2 425 endmember) is -27.4‰, which is close to what would be expected from OM remineralization. For instance, CO2 from remineralized OM in Greenlandic heath soils ranged between approximately -27 to -25‰ (Ravn et al., 2020), and between -20 to -30‰ for thawed Alaskan permafrost soils (Mauritz et al., 2019), both of which may be similar to subglacial organic matter. An additional correlation is observed between CO2-total and NH4 concentrations for Isunnguata samples, which would be expected from OM remineralization (Fig. 8b). The low-CO2 end member could reflect atmospheric CO2 input, which should 430 result in a d 13 C-CO2 value of approximately -8‰. While the d 13 C-CO2 value of the lowest-CO2-total samples in the Isunnguata Keeling plot (e.g. highest [CO2-total] -1 not including the outlier) are slightly depleted compared to atmospheric values at -12.1‰ (Fig. 8a), even the lowest CO2 concentrations measured at Isunnguata are supersaturated with respect to atmospheric concentrations (Fig. 3g). Supersaturation suggests that OM remineralization contributes CO2 even for low CO2-concentration samples and isotopically depletes the subglacial CO2 reservoir. 435 While both atmospheric CO2 (CO2-atm) and CO2 derived from OM remineralization (CO2-OM) provide carbonic acid to drive subglacial mineral weathering as well as CO2 supersaturation in Isunnguata discharge, their absolute and relative contributions are controlled by different processes. Understanding the controls of CO2 acquisition may improve understanding of subglacial carbon dynamics as well as the conditions necessary for subglacial environments to become CO2 sources to the atmosphere. CO2-OM is the dominant source to CO2-total at Isunnguata in the early and late melt seasons (Fig. 9a) when discharge 440 is low (Fig. 9c), while CO2-atm is the dominant CO2 source in the peak melt season when discharge is high. This switch in dominant CO2 sources occurs because the magnitude of CO2-OM has a strong negative association with discharge, approaching zero during maximum discharge times (Fig. 9d), while CO2-atm remains relatively constant over the melt season (Fig. 9b) and the range of discharges (Fig. 9d). The high contributions of CO2-OM during low discharge could reflect higher residence times that allow greater biogeochemical modification and accumulation of OM remineralization reaction products, including CO2. 445 The chemostatic behavior for CO2-atm indicates that invasion of atmospheric CO2 is independent of the extent of chemical weathering, which exhibits strong seasonal variation (Fig. 6a). Chemostatic behavior of CO2-atm may indicate that CO2-OM maintains CO2 concentrations at or above atmospheric saturation concentrations, and no additional atmospheric CO2 dissolution would be needed to maintain equilibrium. Chemostatic behavior of CO2-atm could additionally indicate that the Isunnguata subglacial drainage is largely closed to atmospheric exchange. Both mechanisms are supported by the consistent 450 CO2 supersaturation observed in subglacial discharge at Isunnguata (Fig. 3g), which suggests that limited atmospheric exchange prevents significant outgassing prior to discharge. These results first imply that CO2 supersaturation due to OM remineralization is likely during low flow conditions in systems that are relatively closed to the atmosphere. Moreover, CO2-OM is the main driver for mineral weathering during low-flow conditions: while CO2-MW exceeds -150 µM late in the melt season (Fig. 6a), CO2-atm provides only about a third of this CO2 (Fig.8b), suggesting that the remainder was driven by  In this case, only a fraction of the mineral weathering products measured in subglacial outflow at Isunnguata are directly involved in the sequestration of atmospheric CO2, and the majority of mineral weathering serves to consume CO2 from in situ CO2-OM production.
While Additional subglacial CO2 sources could include atmospheric CO2 contained in ice bubbles, or lithogenic CO2 liberated by mechanical grinding, though both of these sources would be expected to enrich rather than deplete the d 13 C-CO2 values of the samples relative to modern atmospheric d 13 C-CO2 values. Ice bubbles contain gaseous CO2 at concentrations and 470 isotopic compositions reflecting atmospheric conditions during ice formation. While heterogeneity may result from gas bubbles recording changes in atmospheric CO2, variability in d 13 C-CO2 of gas bubble CO2 should be only a few per mil, which is small compared to the variation observed in Russell and Kiattut Sermiat samples (Tipple et al., 2010;Fig. 8a). Gas bubble CO2 should also be 13 C-enriched compared to modern atmospheric CO2 due to fossil fuel contributions, and thus would be unlikely to cause the variation in sample d 13 C-CO2 values that are more 13 C-depleted than modern atmospheric d 13 C-CO2 475 values (Fig. 8a). Recent work has also highlighted the potential for subglacial mechanical grinding to result in CO2 production through the volatilization of fluid inclusions (Macdonald et al., 2018). While volatilization of fluid inclusions through mechanical grinding was found to produce sufficient CO2 to drive approximately 20% of mineral weathering in Svalbard subglacial environments, the expected isotopic composition of lithogenic CO2 is more 13 C-enriched than our measured d 13 C-CO2 values. Because mechanical grinding should produce CO2 with an isotopic composition reflecting the lithogenic source, 480 (Lüders et al., 2012), its contributions here are likely limited. For instance, estimates of d 13 C for bulk hydrocarbons in fluid inclusions in the Ilímaussaq alkaline complex of South Greenland have values of -4.5±1.5‰ (Madsen, 2001), which is close to the d 13 C-CO2 of CO2 in fluid inclusions in the Bamble granulite sector of South Norway (~ -6‰; Newton et al., 1980).
There is an additional possibility of atmospheric exchange between the subglacial outlet site and our water sampling sites that could serve as an additional CO2 source or sink. However, atmospheric CO2 exchange after discharge would have the same 485 impact on Keeling plots as atmospheric CO2 exchange prior to discharge. Although Kiattut Sermiat CO2 concentrations are undersaturated with respect to atmospheric concentrations and would promote invasion of atmospheric CO2, measured d 13 C-CO2 values are more 13 C-depleted than modern atmospheric CO2 and are not consistent with atmospheric CO2 as the sole or dominant source of CO2 to glacial meltwater samples (Fig. 8a). While more information is needed to determine the sources of CO2 to Russell and Kiattut Sermiat samples, d 13 C-CO2 values of samples from both sites imply mixing between a 13 C-depleted 490 CO2 source, such as OM remineralization, and one or more 13 C-enriched CO2 sources, such as atmospheric or lithogenic CO2.

Conclusions
Subglacial reactions impact the concentrations of CO2 and CH4 in subglacial discharge of GrIS, which act as either sources or sinks of GHG to the atmosphere. CH4 concentrations of subglacial discharge are likely controlled by the availability of subglacial OM to drive methanogenesis as well as the extent of CH4 oxidation. Regional differences in subglacial OM 495 deposits may account for the occurrence of methanogenesis in southwest outlet glaciers (Isunnguata and Russell) in this and other studies, contrasting with southern Kiattut Sermiat where little CH4 production occurs. During the early melt season, oxidation consumes nearly 50% of CH4 produced at southwest sites, and relationships between discharge and CH4 oxidation (ec and fox) suggest that CH4 oxidation depends on longer subglacial residence time during periods of low discharge. While mineral weathering consumes CO2 throughout the melt season at all three sites, additional CO2 resupplied from atmospheric 500 and subglacial sources increases the CO2 concentrations of subglacial discharge. The magnitude of additional CO2 sources (CO2-total) is insufficient to maintain atmospheric equilibrium at Kiattut Sermiat, leading subglacial discharge to be a sink of atmospheric CO2, while CO2-total maintains close to atmospheric equilibrium concentrations at the Russell Glacier. At Isunnguata, however, OM remineralization produces more CO2 than is consumed by mineral weathering and causes meltwater to be a source of CO2 to the atmosphere. This finding implies that subglacial mineral weathering serves to partially or fully 505 consume CO2 produced from in situ sources under the GrIS but does not necessarily result in direct consumption of modern atmospheric CO2. The important role of OM remineralization in subglacial environments of the GrIS determined by this and other studies also implies links between subglacial OM deposits and export of other biogeochemical solutes from the GrIS, including nutrients as well as redox-sensitive elements. While the export of nutrients from the GrIS has been the focus of numerous studies (Bhatia et al., 2013;Hawkings et al., 2016;Lawson et al., 2014), little is currently known regarding the role 510 of OM sources in governing these exports. Given the variability in GHG concentrations observed in this study, constraining the extent of heterogeneity in outlet glaciers of the GrIS as well as the biogeochemical, hydrologic, and geologic controls of this heterogeneity will be important for upscaling atmospheric fluxes as well as efforts to predict impacts of ice loss on carbon budgets due to current and future melting of the GrIS.