Archive Meta Data

Author(s)

Jesper Riis Christiansen

Title

Subglacial carbon emissions data June 2018 - May 2019

Description

Data collection related to the manuscript/paper "Carbon emissions from the edge of the Greenland Ice sheet reveal subglacial processes of methane and carbon dioxide turnover".

The data are the basis for alle four figures and 6 supplementary figures in the manuscript/paper. Data was collected in three campaigns in June 2018, August 2018 and May 2019 at the site in Kangerlussuaq at the edge of the glacier Isunnguata Sermia (67°09’16.40’’N 50°04’08.48’’W).
Methods are described below. For correct formatting of equations the reader is referred to the publication:

2.1 Site description
The study site is located at an elevation of 450 m above sea level at a lateral subglacial meltwater outlet on the southern flank at the terminus of the Isunnguata Sermia Glacier at the western margin of the GrIS (67°09’16.40’’N 50°04’08.48’’W). 
The area in front of the meltwater outlet consists of abraded granodioritic gneiss bedrock with large boulders and patches of gravel, sand and silt deposited by meltwater. The glacier front contained highly irregular cracks and air-filled cavities, which changed over the season as the ice melted and deformed (Figure 1). 
The landscape in the Kangerlussuaq area is typical of west Greenland, where numerous, narrow and up to 600 meter deep valleys are oriented in a East - West direction. These valleys extend below the ice sheet, and subglacial valleys can in places reach depths of hundreds of meters below sea level. Deglaciation and re-advance of the GrIS in this region during the Holocene has resulted in buried subglacial carbon rich sediments that were once exposed (Kellerman et al., 2020; Kohler et al., 2017). In the proglacial zone of the study area continuous permafrost extends at places up to 350 meters below the surface (Drake et al., 2017), but the Isunnguata glacier and GrIS in this area is warm based with an annual ice flow of 150-200 meters and surface meltwater reaching the base of the glacier (Graly et al., 2014).
We sampled melt water and gas at a lateral subglacial outlet to the Isunnguata Sermia glacier draining the GrIS in West Greenland (Figure 1 top panel). The sampling was done during three campaigns covering the periods May 3rd to 6th 2019, June 18th to 28th and August 16th to 24th 2018, during which the cross sectional area of the subglacial outlet changed size and position along the ice edge (Figure 1a-c). These periods are assumed to represent the early, middle and late stages of a typical melt season.

2.2 Measurements of subglacial air velocity, temperature, humidity, atmospheric pressure and water level
At the end of an aluminium pole that extended under the ice for retrievel of unmixed subglacial air we attached instrumentation to measure subglacial air velocity (hot-wire anemometer, model 313-T-DCI-F900-L-O, Onset Computer Cooperation, USA), temperature and humidity (model 313-S-THB-M008, Onset Computer Cooperation, USA). The anemometer was positioned so it measured the wind movement perpendicular to the cross section. Atmospheric pressure was measured outside the cave (model 313-S-BPB-CM50, Onset Computer Cooperation, USA). The data were recorded on a HOBO datalogger (model U30-NRC-VIA-05-S100-000, Onset Computer Cooperation, USA) at 10 second intervals. These measurements were conducted during the June and August campaigns only. We were only able to measure the air velocity for a short period in June as the sensor was damaged by water spray in the ice cave.
During the August 2018 campaign, we also installed an underwater pressure transducer (Onset Computer Corporation, USA) in the outlet stream to estimate the temporal variability of the water level. Air pressure from the meteorological station was used as the atmospheric reference needed to calculate the water level above the pressure transducer. The water level was assumed as a proxy for melt water runoff, but the discharge volume was not estimated.

2.3 Measurements of gaseous subglacial CH4 and CO2 mole fractions and flux calculation
Dry mole fractions of CH4 and CO2 in the subglacial air were measured with a portable CH4/CO2/H2O analyzer (Ultraportable Greenhouse Gas Analyzer (UGGA), ABB Los Gatos Research, USA) powered by a 12 V 100 Ah LiFePO4 battery. Due to shifting positions and geometries of the subglacial cave, the gas sampling setup with the UGGA was not identical during all campaigns, but generally followed the same procedure (Figure 1a-c). The cross sectional areas of the outlet during the three campaigns were estimated based on field observations of the dimensions (height and width) of the opening (Figure 1a-c). Gas measurements were performed by attaching a tube to a 9 m aluminium pole and sampling the air inside the subglacial cavities (Figure 1a-c). A water trap fixed to the end of the aluminium pole ensured a liquid free air stream to the gas analyzer. 
The net CH4 and CO2 emission (g CH4 s-1 or g CO2 s-1) across the entire cross sectional area from the subglacial cave to the atmosphere was calculated as a mass flow of air through the estimated cross section area according to equation 1:
F_(〖CO〗_2/〖CH〗_4 )=C*ῡ*A*273.15/((M_v*T_a ) )*M*〖10〗^(-6) 						(equation 1)
Where C is the measured 0.1 Hz dry mole fraction (μmol mol-1) of CO2 or CH4, ῡ is the wind speed (m s-1) measured every 10 seconds perpendicular to the cross section, A is the cross sectional area at the given measurement period (m2), Mv is the molar volume (m3 mol-1), Ta is the air temperature (°K) measured in the cavity, M is the molar mass of CO2 or CH4 (g mol-1), the constant 10-6 converts the flux from µg to g CO2/CH4. The cross sectional area was estimated based on the width and height measured in the field (Figure 1a-c). To estimate and compare the net CH4 and CO2 emission between campaigns we assumed that the average wind speed (0.8±0.28 m s-1) measured in June 2018 and air temperature (0.2°C) in the cavity was similar between and constant during the three measurement periods. The average hourly net emission for each measurement period was then calculated as the sum of 0.1 Hz emissions over the measurement period divided by length in hours of the measurement periods. This approach provide at best a rough and uncertain estimate, referred to as “plausible range”, and was calculated as the emission for the minimum wind speed at the minimum cross section area and maximum wind speed for the maximum cross section area. The impact of short term influx of CH4 and CO2 from the atmosphere to the cave, due to turbulent mixing, was accounted for by averaging the 0.1 Hz effluxes over the measurement period.

2.4 Collection of discrete water and gas samples
Water and gas samples were taken at three different locations after the subglacial water and air had mixed to different degrees with the ambient environment. For the air samples, the simultaneous variations in mole fraction and isotopic composition were used to determine the isotopic composition of the source (δ13C-CH4, δ2H-CH4 and δ13C-CO2) of the subglacial CH4 and CO2 using the Keeling plot approach. This is a widely used method to determine the isotope composition of unknown sources of CO2 or CH4 in situations where CH4 or CO2 from a source (in our case the subglacial environment) is added to a constant background (atmosphere) (Pataki et al., 2003).
Water and gas were sampled twice per day, in the morning and evening, assumed to represent low and high water flow derived from the water level measurements. In 2018, samples were gathered during the periods 22nd – 26th June and 19th – 22nd August. 
Air samples were collected in 2L gas tight aluminium foil bags (Supel™-Inert Multi-Layer Foil, Sigma-Aldrich, USA) which were filled by a small diaphragm pump. We sampled gas from three locations (Figure 1a-c); inside the ice cave, representing the least mixed subglacial air we could possibly sample (minimal mixing with atmosphere), right outside the ice cave (subglacial air mixed with atmospheric air) and 2 km from the ice edge (background atmosphere, no subglacial air signal).
For practical reasons the water was sampled at slightly different positions than the gas. Thus, the first water sample representing the subglacial water was sampled right where the meltwater exists the ice (PW1), the second sample (PW2) 200 meter downstream and the third sample was taken at the same position as the third gas sample, 2 km away from the ice edge (PW3). Unfiltered water was sampled in 120 mL glass bottles with butyl rubber septa and tightened with aluminium screw caps. The bottles were rinsed three times with melt water and filled under water ensuring that no bubbles were included. Immediately after sampling, 12 μL saturated HgCl2 solution was added to the bottles to exclude further biological activity (Magen et al., 2014). Water was sampled in duplicates, one sample for measurement of dissolved CH4 and another for measurement of CH4 isotopic composition. 
Gas and water samples were stored cold and dark until analysis, except during transport from Greenland to Denmark where samples were transported in the cargo hold of the airplane. Transport resulted in loss of three gas samples, but water samples remained intact. Upon arrival in Denmark the gas bags were immediately sent to Utrecht over land and transferred to glass bottles for longer term storage until isotopic analyses were possible. The total time from sampling to extraction was up to 14 days.

2.5 Dissolved CH4 concentrations
The dissolved CH4 was extracted using headspace mixing and the concentration was calculated according to the method outlined in Magen et al. (2014). Shortly, 10 mL of water (VHS) was replaced with CH4 free N2 gas and the headspace was afterwards pressurized to 2 atmosphere (PHS), by adding another 10 mL N2 amounting to 20 mL of gas in the headspace (Vgas). The sample was then thoroughly stirred on a shaking table with 150 RPM for three minutes. A 5 mL gas sample was retrieved by syringe from the headspace and transferred to an evacuated 3 mL exetainer with a butyl rubber screw cap (Labco, UK). The pressurization of the exetainer was done to facilitate subsequent gas chromatography analysis. The CH4 mole fraction in the headspace (CH4,mf) of extracted gas samples was determined on a gas chromatograph equipped with an FID detector. CH4 was separated on a HayeSep Q column heated to 60°C, with pure N2 5.0 as carrier gas. Using a five-point calibration curve the headspace CH4 mole fraction in ppm was determined. The total dissolved CH4 was calculated as the sum of the headspace CH4 and CH4 still dissolved in the water after shaking (Magen et al. 2014). The ideal gas law was used (laboratory temperature at extraction was 23°C) to convert the headspace concentration to gas amount (mole) (equation 2). The dissolved CH4 in the remaining 110 mL water was calculated by multiplying the Bunsen coefficient for 0°C (water temperature at sampling) at zero salinity (assumed as we have no data) with the amount of headspace CH4 to calculate the remaining dissolved CH4 in water (Yamamoto et al., 1976), accounting for the ratio of water and gas volume (Magen et al., 2014) (equation 3). 
〖CH〗_(4,HS)=〖CH〗_(4,mf)* V_HS*P_HS/(R*T_HS )  [μmol L^(-1)] 					(equation 2)
〖CH〗_(4,water)=β*(〖CH〗_(4,conc)*V_gas*V_water/V_HS )/(R*T_water )  [μmol L^(-1)] 					(equation 3)
Where CH4,conc is the headspace CH4 mole fraction in ppm, VHS is the headspace volume in L, PHS is the headspace pressure in atm, R is the gas constant (atm L K-1 mol-1), THS is the headspace temperature in °K, β is the Bunsen coefficient, Vgas is the total volume of gas in headspace in L, Vwater is the water volume after replacement in L, Twater is the water temperature (similar to THS).

2.6 Dissolved CO2 concentrations
Dissolved CO2 in meltwater was measured in situ using an eosGP2 probe (Eosense Inc., Canada) connected to a Campbell CR1000 datalogger (Campbell Scientific Inc., USA) during the June 2018 campaign. The sampling interval was 10 seconds and dissolved CO2 concentrations given in ppm. A custom calibration for measurements at CO2 concentrations close to the atmospheric equilibrium had been done prior to the field work by Eosense. Before each deployment, we let the eosGP2 probe equilibrate with the atmospheric background CO2 concentration for approximately one hour to monitor possible drift and/or sensitivity of the response of the CO2 signal when switching the probe between the aqueous and gaseous phases. At deployment the eosGP2 probes were fixed in place and the diffusion membrane initially placed 15 cm below the surface of the meltwater at low flow conditions. 

2.7 Isotopic analyses of gas and water samples
The isotopic composition of CH4 (δ13C-CH4, δ2H-CH4) was measured using continuous-flow isotope ratio mass spectrometry (CF-IRMS) on a ThermoFinnigan Deltaplus XL isotope ratio mass spectrometer. The air samples were injected via a mass flow controller into the sample loop of the extraction system and further processed and analyzed as described in Röckmann et al. (2016). The CH4 in the water samples was extracted with a headspace mixing method and further analyzed on the same analytical system, as described in Jacques et al. (2020). Further information on the data processing is available in Brass and Röckmann (2010) and Sapart et al. (2011). Specifically, the CH4 isotopic data were corrected to account for system variability and non-linearity effects and reported in ‰ vs VPDB for δ13C values and ‰ vs VSMOW for δ2H values. The measurement reproducibility was calculated from the standard deviation of reference air injections.
The isotopic composition of CO2 (δ13C and δ18O) was analyzed with the CF-IRMS system described in Naus et al. (2018) and Pathirana et al. (2015). This system is primarily meant for CO isotopes, but can also analyze CO2 isotopes is small samples (~ 2 ml air at normal atmospheric mole fractions). In short, the CO2 is cryogenically separated from the air, further purified chromatographically, and then injected into the IRMS via an open split inlet. The results are related to the VPDB and VSMOW scales via a reference air cylinder with known isotopic composition. The typical precision, estimated as repeatability of multiple measurements of a constant gas (Target cylinder), is about 3 ppm for the CO2 mole fractions, and 0.05 ‰ and 0.14 ‰ for δ13C and δ18O respectively.

Cited literature
Brass, M., & Röeckmann, T. (2010). Continuous-flow isotope ratio mass spectrometry method for carbon and hydrogen isotope measurements on atmospheric methane. Atmospheric Measurement Techniques. https://doi.org/10.5194/amt-3-1707-2010
Drake, H., Suksi, J., Tullborg, E. L., & Lahaye, Y. (2017). Quaternary redox transitions in deep crystalline rock fractures at the western margin of the Greenland ice sheet. Applied Geochemistry, 76, 196–209. https://doi.org/10.1016/j.apgeochem.2016.12.001
Graly, J. A., Humphrey, N. F., Landowski, C. M., & Harper, J. T. (2014). Chemical weathering under the Greenland Ice Sheet. Geology, 42(6), 551–554. https://doi.org/10.1130/G35370.1
Jacques, C., Gkritzalis, T., Tison, J.-L., Hartley, T., van der Veen, C., Röckmann, T., et al. (2020). Carbon and Hydrogen Isotope Signatures of Dissolved Methane in the Scheldt Estuary. Estuaries and Coasts. https://doi.org/10.1007/s12237-020-00768-3
Kellerman, A. M., Hawkings, J. R., Wadham, J. L., Kohler, T. J., Stibal, M., Grater, E., et al. (2020). Glacier outflow dissolved organic matter as a window into seasonally changing carbon sources: Leverett Glacier, Greenland. Journal of Geophysical Research: Biogeosciences, 1–16. https://doi.org/10.1029/2019jg005161
Kohler, T. J., Žárský, J. D., Yde, J. C., Lamarche-Gagnon, G., Hawkings, J. R., Tedstone, A. J., et al. (2017). Carbon dating reveals a seasonal progression in the source of particulate organic carbon exported from the Greenland Ice Sheet. Geophysical Research Letters, 44(12), 6209–6217. https://doi.org/10.1002/2017GL073219
Magen, C., Lapham, L. L., Pohlman, J. W., Marshall, K., Bosman, S., Casso, M., & Chanton, J. P. (2014). A simple headspace equilibration method for measuring dissolved methane. Limnology and Oceanography: Methods, 12(9), 637–650. https://doi.org/10.4319/lom.2014.12.637
Naus, S., Röckmann, T., & Popa, M. E. (2018). The isotopic composition of CO in vehicle exhaust. Atmospheric Environment, 177, 132–142. https://doi.org/10.1016/j.atmosenv.2018.01.015
Pataki, D. E., Ehleringer, J. R., Flanagan, L. B., Yakir, D., Bowling, D. R., Still, C. J., et al. (2003). The application and interpretation of Keeling plots in terrestrial carbon cycle research. Global Biogeochemical Cycles, 17(1). https://doi.org/10.1029/2001GB001850
Pathirana, S. L., van der Veen, C., Popa, M. E., & Röckmann, T. (2015). An analytical system for stable isotope analysis on carbon monoxide using continuous-flow isotope-ratio mass spectrometry. Atmospheric Measurement Techniques, 8(12), 5315–5324. https://doi.org/10.5194/amt-8-5315-2015
Röckmann, T., Eyer, S., van der Veen, C., Popa, M. E., Tuzson, B., Monteil, G., et al. (2016). In situ observations of the isotopic composition of methane at the Cabauw tall tower site. Atmospheric Chemistry and Physics, 16(16), 10469–10487. https://doi.org/10.5194/acp-16-10469-2016
Sapart, C. J., van der Veen, C., Vigano, I., Brass, M., van de Wal, R. S. W., Bock, M., et al. (2011). Simultaneous stable isotope analysis of methane and nitrous oxide on ice core samples. Atmospheric Measurement Techniques, 4(12), 2607–2618. https://doi.org/10.5194/amt-4-2607-2011
Yamamoto, S., Alcauskas, J. B., & Crozier, T. E. (1976). Solubility of methane in distilled water and seawater. Journal of Chemical & Engineering Data, 21(1), 78–80. https://doi.org/10.1021/je60068a029