Northern lakes are considered a major source of atmospheric methane (CH4), a potent GHG1,2. However, large uncertainties in their emissions (7–26 Tg CH4 yr–1; ref. 2) arise from challenges in upscaling field data, including fluxes by ebullition (bubbling), the dominant emission pathway2. Remote sensing of ebullition would allow detailed mapping of regional emissions but has hitherto not been developed. Here, we show that lake ebullition can be imaged using synthetic aperture radar remote sensing during ice-cover periods by exploiting the effect of ebullition on the texture of the ice–water interface. Applying this method to five Alaska regions and combining spatial remote sensing information with year-round bubble-trap flux measurements, we create ebullition-flux maps for 5,143 Alaskan lakes. Regional lake CH4 emissions, based on satellite remote sensing analyses, were lower compared to previous estimates based on upscaling from individual lakes2,3 and were consistent with independent airborne CH4 observations. Thermokarst lakes formed by thaw of organic-rich permafrost had the highest fluxes, although lake density and lake size distributions also controlled regional emissions. This new remote sensing approach offers an opportunity to improve knowledge about Arctic CH4 fluxes and helps to explain long-standing discrepancies between estimates of CH4 emissions from atmospheric measurements and data upscaled from individual lakes.
Subscribe to Journal
Get full journal access for 1 year
only $4.92 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Source data for Figs. 1 and 3 and Extended Data Figs. 3–9 are presented with the paper. SAR data that support the findings of this study were used under license for this study and are available from the Alaska Satellite Facility (www.asf.alaska.edu) for non-commercial users. The SAR-based CH4 lake shapefiles generated during this study are available at NASA’s Distributed Active Archive Center for Biogeochemical Dynamics at the Oak Ridge National Laboratory (ONRL DAAC): 10.3334/ORNLDAAC/1790 (ref. 55).
Bastviken, D., Tranvik, L. J., Downing, J. A., Crill, P. M. & Enrich-Prast, A. Freshwater methane emissions offset the continental carbon sink. Science 331, 50 (2011).
Wik, M., Varner, R. K., Anthony, K. W., MacIntyre, S. & Bastviken, D. Climate-sensitive northern lakes and ponds are critical components of methane release. Nat. Geosci. 9, 99–105 (2016).
Walter Anthony, K. M., Anthony, P., Grosse, G. & Chanton, J. Geologic methane seeps along boundaries of Arctic permafrost thaw and melting glaciers. Nat. Geosci. 5, 419–426 (2012).
McGuire, A. D. et al. Sensitivity of the carbon cycle in the Arctic to climate change. Ecol. Monogr. 79, 523–555 (2009).
AMAP Assessment 2015: Methane as an Arctic Climate Forcer (AMAP, 2015).
Bruhwiler, L. et al. CarbonTracker-CH4: an assimilation system for estimating emissions of atmospheric methane. Atmos. Chem. Phys. 14, 8269–8293 (2014).
Rigby, M. et al. Role of atmospheric oxidation in recent methane growth. Proc. Natl Acad. Sci. USA 114, 5373–5377 (2017).
Thornton, B. F., Wik, M. & Crill, P. M. Double-counting challenges the accuracy of high-latitude methane inventories. Geophys. Res. Lett. 43, 12569–12577 (2016).
Jørgensen, C. J., Lund, K. M. J., Westergaard‐Nielsen, A. & Elberling, B. Net regional methane sink in High Arctic soils of northeast Greenland. Nat. Geosci. 8, 20–23 (2015).
Peltola, O. et al. Monthly gridded data product of northern wetland methane emissions based on upscaling eddy covariance observations. Earth Syst. Sci. Data 11, 1263–1289 (2019).
Mastepanov, M. et al. Large tundra methane burst during onset of freezing. Nature 456, 628–630 (2008).
Wik, M., Crill, P. M., Varner, R. K. & Bastviken, D. Multiyear measurements of ebullitive methane flux from three subarctic lakes. J. Geophys. Res. Biogeosci. 118, 1307–1321 (2013).
Walter Anthony, K. M. & Anthony, P. Constraining spatial variability of methane ebullition seeps in thermokarst lakes using point process models. J. Geophys. Res. Biogeosci. 118, 1015–1034 (2013).
Lehner, B. & Dőll, P. Development and validation of a global database of lakes, reservoirs and wetlands. J. Hydrol. 296, 1–22 (2004).
Paltan, H., Dash, J. & Edwards, M. A refined mapping of Arctic lakes using Landsat imagery. Int. J. Remote Sens. 36, 5970–5982 (2015).
Lindgren, P. R., Grosse, G., Walter Anthony, K. M. & Meyer, F. J. Detection and spatiotemporal analysis of methane ebullition on thermokarst lake ice using high-resolution optical aerial imagery. Biogeosciences 13, 27–44 (2016).
Engram, M., Anthony, K. W., Meyer, F. J. & Grosse, G. Synthetic aperture radar (SAR) backscatter response from methane ebullition bubbles trapped by thermokarst lake ice. Can. J. Remote Sens. 38, 667–682 (2013).
Walter Anthony, K. M. et al. Estimating methane emissions from northern lakes using ice-bubble surveys. Limnol. Oceanogr. Methods 8, 592–609 (2010).
Rasilo, T., Prairie, Y. T. & del Giorgio, P. A. Large-scale patterns in summer diffusive CH4 fluxes across boreal lakes, and contribution to diffusive C emissions. Glob. Change Biol. 21, 1124–1139 (2015).
Bastviken, D., Cole, J., Pace, M. & Tranvik, L. Methane emissions from lakes: dependence of lake characteristics, two regional assessments, and a global estimate. Glob. Biogeochem. Cycles 18, GB4009 (2004).
Kessler, M. A., Plug, L. J. & Walter Anthony, K. M. Simulating the decadal- to millennial-scale dynamics of morphology and sequestered carbon mobilization of two thermokarst lakes in NW Alaska. J. Geophys. Res. 117, G00M06 (2012).
Walter Anthony, K. et al. Methane emissions proportional to permafrost carbon thawed in Arctic lakes since the 1950s. Nat. Geosci. 9, 679–682 (2016).
Walter Anthony, K. et al. 21st-century modeled permafrost carbon emissions accelerated by abrupt thaw beneath lakes. Nat. Commun. 9, 3262 (2018).
Aben, R. C. H. et al. Cross continental increase in methane ebullition under climate change. Nat. Commun. 8, 1682 (2017).
Nitze, I., Grosse, G., Jones, B. M., Romanovsky, V. E. & Boike, J. Remote sensing quantifies widespread abundance of permafrost region disturbances across the Arctic and Subarctic. Nat. Commun. 9, 5423 (2018).
Saunois, M. et al. The global methane budget: 2000–2012. Earth Syst. Sci. Data Discuss. 8, 54–81 (2016).
Arp, C. D. et al. Threshold sensitivity of shallow Arctic lakes and sublake permafrost to changing winter climate. Geophys. Res. Lett. 43, 6358–6365 (2016).
Engram, M., Arp, C. D., Jones, B. M., Ajadi, O. A. & Meyer, F. J. Analyzing floating and bedfast lake ice regimes across Arctic Alaska using 25 years of space-borne SAR imagery. Remote Sens. Environ. 209, 660–676 (2018).
Surdu, C. M., Duguay, C. R., Brown, L. C. & Fernández Prieto, D. Response of ice cover on shallow lakes of the North Slope of Alaska to contemporary climate conditions (1950–2011): radar remote-sensing and numerical modeling data analysis. Cryosphere 8, 167–180 (2014).
DelSontro, T., Boutet, L., St-Pierre, A., del Giorgio, P. A. & Prairie, Y. T. Methane ebullition and diffusion from northern ponds and lakes regulated by the interaction between temperature and system productivity. Limnol. Oceanogr. 61, S62–S77 (2016).
Weeks, W. F., Fountain, A. G., Bryan, M. L. & Elachi, C. Differences in radar return from ice-covered North Slope lakes. J. Geophys. Res. 83, 4069–4073 (1978).
Jeffries, M. O., Morris, K., Weeks, W. F. & Wakabayashi, H. Structural and stratigraphic features and ERS-1 synthetic-aperture radar backscatter characteristics of ice growing on shallow lakes in NW Alaska, winter 1991–1992. J. Geophys. Res. Oceans 99, 22459–22471 (1994).
Hall, D. K., Fagre, D. B., Klasner, F., Linebaugh, G. & Liston, G. E. Analysis of ERS-1 synthetic-aperture-radar data of frozen lakes in northern Montana and implications for climate studies. J. Geophys. Res. Oceans 99, 22473–22482 (1994).
Mellor, J. Bathymetry of Alaskan Arctic Lakes: A Key to Resource Inventory with Remote Sensing Methods. PhD thesis, Institute of Marine Science, University of Alaska Fairbanks (1982).
Duguay, C. R., Pultz, T. J., Lafleur, P. M. & Drai, D. RADARSAT backscatter characteristics of ice growing on shallow sub-Arctic lakes, Churchill, Manitoba, Canada. Hydrol. Process. 16, 1631–1644 (2002).
Walter, K. M., Engram, M., Duguay, C. R., Jeffries, M. O. & Chapin, F. S. The potential use of synthetic aperture radar for estimating methane ebullition from Arctic lake. J. Am. Water Resour. Assoc. 44, 305–315 (2008).
Lee, J.-S. & Pottier, E. in Polarimetric Radar Imaging: From Basics to Applications Vol. 1 (eds Lee, J.-S. & Pottier, E.) 265–299 (CRC Press Taylor & Francis Group, 2009).
Leconte, R. et al. A controlled experiment to retrieve freshwater ice characteristics from an FM-CW radar system. Cold Reg. Sci. Technol. 55, 212–220 (2009).
Scandella, B. P., Varadharajan, C., Hemond, H. F., Ruppel, C. & Juanes, R. A conduit dilation model of methane venting from lake sediments. Geophys. Res. Lett. 38, L06408 (2011).
Greene, S., Walter Anthony, K. M., Archer, D., Sepulveda-Jauregui, A. & Martinez-Cruz, K. Modeling the impediment of methane ebullition bubbles by seasonal lake ice. Biogeosciences 11, 6791–6811 (2014).
Walter, K. M., Zimov, S. A., Chanton, J. P., Verbyla, D. & Chapin, F. S. Methane bubbling from Siberian thaw lakes as a positive feedback to climate warming. Nature 443, 71–75 (2006).
Walter, K. M., Chanton, J. P., Chapin, F. S., Schuur, E. A. G. & Zimov, S. A. Methane production and bubble emissions from Arctic lakes: isotopic implications for source pathways and ages. J. Geophys. Res. Biogeosci. 113, G00A08 (2008).
Zimov, S. A. et al. in Permafrost Response on Economic Development, Environmental Security and Natural Resources, NATO Science Series, Vol. 76 (eds Paepe, R. & Melnikov, V.) 511–524 (Springer, 2001).
Engram, M., Anthony, K. W., Meyer, F. J. & Grosse, G. Characterization of L-band synthetic aperture radar (SAR) backscatter from floating and grounded thermokarst lake ice in Arctic Alaska. Cryosphere 7, 1741–1752 (2013).
Serafimovich, A. et al. Upscaling surface energy fluxes over the North Slope of Alaska using airborne eddy-covariance measurements and environmental response functions. Atmos. Chem. Phys. 18, 10007–10023 (2018).
Hartmann, J., Gehrmann, M., Kohnert, K., Metzger, S. & Sachs, T. New calibration procedures for airborne turbulence measurements and accuracy of the methane fluxes during the AirMeth campaigns. Atmos. Meas. Tech. 11, 4567–4581 (2018).
Aubinet., M., Vesala, T. & Papale, D. Eddy Covariance: A Practical Guide to Measurement and Data Analysis (Springer Science & Business Media, 2012).
Foken, T. & Napo, C. J. Micrometeorology (Springer, 2008).
Metzger, S. et al. Spatially explicit regionalization of airborne flux measurements using environmental response functions. Biogeosciences 10, 2193–2217 (2013).
Kljun, N., Calanca, P., Rotach, M. W. & Schmid, H. P. A simple parameterisation for flux footprint predictions. Bound. Layer Meteorol. 112, 503–523 (2004).
Metzger, S. et al. Eddy-covariance flux measurements with a weight-shift microlight aircraft. Atmos. Meas. Tech. 5, 1699–1717 (2012).
Elith, J., Leathwick, J. R. & Hastie, T. A working guide to boosted regression trees. J. Anim. Ecol. 77, 802–813 (2008).
Skamarock, W. C. et al. A Description of the Advanced Research WRF Version 3 NCAR Technical Note NCAR/TN-475+STR (UCAR, 2008).
Jones, B. M. & Grosse, G. Western Arctic Coastal Plain, Lakes and Drainage Gradients (Arctic Landscape Conservation Cooperative, 2013); http://arcticlcc.org/products/spatial-data/show/western-arctic-coastal-plain-lakes-and-drainage-gradients
Engram, M., Walter Anthony, K. M., & Meyer, F.J. SAR-based Lake Ebullition Estimates for Five Alaska Regions (Oak Ridge National Laboratory NASA Distributed Active Archive Center, 2020); https://doi.org/10.3334/ORNLDAAC/1790
Metzger, S. et al. eddy4R 0.2.0: a DevOps model for community-extensible processing and analysis of eddy-covariance data based on R, Git, Docker, and HDF5. Geosci. Model Dev. 10, 3189–3206 (2017).
We thank S. Metzger for processing the airborne EC raw data and for advice on the machine-learning steps in creating the regional airborne EC flux map using the free eddy4R software framework. We thank the Alaska Satellite Facility for SAR data and processing tools, and the State of Alaska Division of Geological and Geophysical Surveys for high-resolution elevation data. We thank A. Bondurant, P. Anthony, D. Whiteman, L. Plug, C. Arp, B. Jones, L. Farquharson, A. Strohm, L. Oxtoby, M. Geai, L. McFadden, M. Smith and P. Lindgren for assistance in collecting field data. This study was part of the Arctic-Boreal Vulnerability Experiment and funded by NASA (grant nos. NNX15AU49A and NNN12AA01C) and National Science Foundation ARCSS grant no. 1500931. Airborne work was supported by the Helmholtz Association of German Research Centres through a Helmholtz Young Investigators Group grant to T.S. (grant no. VH-NG-821) and is a contribution to the Helmholtz Climate Initiative REKLIM (Regional Climate Change). The AirMeth campaigns were mainly funded by Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research. G.G. was additionally supported by ERC grant no. 338335.
The authors declare no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Right hand panels are a focused view on areas of ice blocks shown in left hand panels. Arrows point to the top of the ice blocks. Thick black marks on the metre stick indicate 10 cm. Panels a and c show ice blocks harvested from boreal lakes with strong ebullition seeps, with close-up view of rough ice bottom (~ 4 cm deep) shown in panels b and d. Panel e and close-up f show an ice block harvested from an ebullition-free gravel pit pond where artificial ebullition was conducted as described in Engram et al.44, demonstrating cavities below gas bubbles are caused by uneven ice growth due to the thermal insulation of gas bubbles. Panel g shows a ‘wedding cake’ pattern of stacked bubbles that had partly vented to the atmosphere. In h, fingers shown for scale point to a crack in the ice, through which gas from the ice pocket periodically escaped. This wedding cake feature is depicted in Extended Data Fig. 2e.
The main roughness phenomenon is caused by insulating bubbles leading to slower ice growth directly beneath the bubble column. The resulting upward cavities in the bottom of ice are filled by water (feature a), partly filled by gas (feature b), or completely filled by gas (feature c). Rough surfaces are shown by black arrows; arrow width depicts expected roughness intensity. Feature d shows a pause in ebullition (a weaker seep) that allows ice growth to ‘catch up’ with surrounding ice sheet. Feature e shows stacked bubbles from a strong seep in a ‘wedding cake’ formation. In this example, the feature has vented the gas to the atmosphere, resulting in multiple ridges and a near-vertical ice–water interface with strong SAR backscatter. Wedding cake cavities typically re-fill with gas and vent again in an alternating cycle. Real portions of dielectric constants for different media are represented by ɛ′.
Extended Data Fig. 3 L-band T11 SAR backscatter versus gas flux from ebullition as measured by field work.
Triangles are 48 training lakes from five geographic regions in Alaska, representing boreal and tundra ecosystems, yedoma and non-yedoma soils, thermokarst, kettle, and anthropogenic lakes resulting from gold or gravel extraction. The regression model for these lakes (backscatter = 1.0*10−3x + 0.019, r2 = 0.91, P < 0.01, n = 48) was inverted (Fig. 1b) to estimate ebullition in individual lakes and regional-scale lake maps. Dotted lines delineate 95% confidence interval. Circles are 16 validation lakes where additional field work was performed in 2016 and 2017 in two geographic regions in Alaska, showing good agreement with the model, but not included in the model. Source data
We show the frequency distributions of the lake-area-weighted mean ebullition fluxes of lakes randomly and repeatedly selected (with no replacement) from all lakes in the SAR scene. We iteratively selected the same number of lakes (n) that we had field data from for our study regions: (a) Barrow Peninsula, n = 5; (b) Atqasuk, n = 7; (c) Toolik n = 9; (d) northern Seward Peninsula n = 11; (e) Fairbanks Yedoma n = 13; (f) Fairbanks Anthropogenic, n = 3 from all lakes in each region, then calculated the regional flux based on the n lakes’ weighted mean SAR values. This approach revealed large variability in the regional lake ebullition estimates when using a small sample size of lakes. Data were binned in 0.5 L gas m−2 yr−1 bins. The blue star represents the field-based flux estimated through field work conducted on n specific lakes. The yellow star indicates the SAR-based flux from all lakes in the SAR scene (see Table 1). Note logarithmic scale for y-axis for the Atqasuk region. Source data
Extended Data Fig. 5 Comparison of SAR-based (yellow) and field-based (blue) fluxes, weighted by area of each lake, alongside higher non-weighted field-based fluxes (red).
Box-plot centre line indicates median; box limits show upper and lower quartiles; whiskers show 1.5 x interquartile range, circles are outliers and asterisks are far-outliers. The scale in panel a omits four far-outliers (asterisks) for the northern Seward Peninsula (at 132, 133, 162, and 845 g CH4 m−2 yr−1) and one far-outlier for the Fairbanks Yedoma region (at 201 g CH4 m−2 yr−1); no near-outliers (open circles) were truncated. A subset of the data for Barrow Peninsula, Atqasuk and Toolik regions are re-projected on a finer scale in panel b. Source data
Extended Data Fig. 6 Difference in airborne EC and SAR mean fluxes as a function of airborne EC coverage of a lake.
Circles represent individual lakes in the Barrow Peninsula region (a) and Atqasuk region (b). Difference values greater than zero (black line) indicate how much higher the airborne EC estimated flux for each lake was compared to the June–July SAR-estimated flux. Difference values lower than zero indicate SAR-based estimates higher than airborne EC-based estimates. All differences are plotted against the areal percentage of each lake covered with airborne EC data. Source data
Extended Data Fig. 7 Difference in airborne EC mean flux and SAR mean flux as a function of lake size.
Circles represent individual lakes in the Barrow Peninsula region (a) and Atqasuk region (b). Difference values greater than zero (black line) indicate how much higher the airborne EC estimated flux for each lake was compared to the June/July SAR-estimated flux. Difference values lower than zero indicate SAR-based estimates higher than airborne EC-based estimates. Note different scale in x-axis for each region. Source data
Three lakes (one each small, medium, large) were randomly selected for each region (Supplementary Discussion), then 25% of the pixels were randomly and stepwise removed to determine the effect of less airborne coverage on the whole-lake flux calculation using the Kruskal-Wallis test. Box-plot centre line indicates median; box limits show upper and lower quartiles; whiskers show 1.5 x interquartile range, dots are outliers. The decreased number of pixels did not lead to significantly different fluxes (alpha = 0.05) in either the Atqasuk region (a) or the Barrow Peninsula region (b). Source data
Percent of contribution of lakes by SAR-based volumetric flux in L gas m−2 yr−1. Low emitting lakes contribute the majority of gas in all tundra regions and in Fairbanks Anthropogenic lakes, while low emitting lakes contribute < 20% in boreal lakes (Tanana River Flats and Fairbanks Yedoma). Source data
Panel a shows within-lake variability in backscatter values for a northern Seward Peninsula lakes in L-band T11 data acquired on November 14, 2009. Within-lake patterns of ebullition detected by SAR in panel a are also visible in high-resolution optical remote sensing data (Geoeye 1 10/21/2009) shown in panels b and c, in which white patches within-lake correspond to ebullition bubbles trapped in lake ice. White outlines i and iii show areas of high ebullition, both in SAR and in Geoeye; white outline ii shows area of low ebullition in both sensors16.
Source data for graph in Fig. 1b.
Source data for graphs.
Source data for graph of empirical regression model.
Source data for histograms.
Source data for box-plots.
Source data for graphs.
Source data for graphs.
Statistical source data for box-plots.
Source data for graph.
About this article
Cite this article
Engram, M., Walter Anthony, K.M., Sachs, T. et al. Remote sensing northern lake methane ebullition. Nat. Clim. Chang. 10, 511–517 (2020). https://doi.org/10.1038/s41558-020-0762-8
Mapping potential signs of gas emissions in ice of Lake Neyto, Yamal, Russia, using synthetic aperture radar and multispectral remote sensing data
The Cryosphere (2021)
Remote Sensing (2021)
Hydrology and Earth System Sciences (2020)