Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Oct 8;10(1):4584.
doi: 10.1038/s41467-019-12541-7.

Global ocean methane emissions dominated by shallow coastal waters

Thomas Weber  1 , Nicola A Wiseman  2   3 , Annette Kock  4
Affiliations

Global ocean methane emissions dominated by shallow coastal waters

Thomas Weber et al. Nat Commun. .

Abstract

Oceanic emissions represent a highly uncertain term in the natural atmospheric methane (CH4) budget, due to the sparse sampling of dissolved CH4 in the marine environment. Here we overcome this limitation by training machine-learning models to map the surface distribution of methane disequilibrium (∆CH4). Our approach yields a global diffusive CH4 flux of 2-6TgCH4yr-1 from the ocean to the atmosphere, after propagating uncertainties in ∆CH4 and gas transfer velocity. Combined with constraints on bubble-driven ebullitive fluxes, we place total oceanic CH4 emissions between 6-12TgCH4yr-1, narrowing the range adopted by recent atmospheric budgets (5-25TgCH4yr-1) by a factor of three. The global flux is dominated by shallow near-shore environments, where CH4 released from the seafloor can escape to the atmosphere before oxidation. In the open ocean, our models reveal a significant relationship between ∆CH4 and primary production that is consistent with hypothesized pathways of in situ methane production during organic matter cycling.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
Global ∆CH4 climatology. a Annual-mean ∆CH4, computed after binning all data into 0.25 × 0.25 monthly climatology. Data points are drawn larger than the grid cells for clarity. b Probability distributions of observed ∆CH4, grouped into four bathymetric regions (see also Supplementary Fig. 2). Boxes span the interquartile range, with black line at median. Black diamonds are mean values, and whiskers span the 5–95th percentiles. Number of datapoints (n) and data density per 109 m2 (N) after binning are listed
Fig. 2
Fig. 2
Machine-learning mapping of ∆CH4. a Annual mean ∆CH4 averaged across an ensemble of 100,000 individual maps generated by the artificial neural network (ANN) method. b Same as a but from random regression forest (RRF) method. c Taylor diagram summarizing the fit of a subset of 100 randomly selected ANN and RRF models to observed ∆CH4, after transformation (see the “Methods” section). Correlation coefficient (R) is shown on the outer angular axis, centered root-mean-squared difference is given by radial distance from REF point, and standard deviation (s.d.) normalized by observed s.d. is the radial distance from the origin (points on the 1.0 line have the same s.d. as observations). ANN and RRF dramatically outperform linear regression and multiple linear regression models by all three metrics
Fig. 3
Fig. 3
Diffusive ocean–atmosphere CH4 flux. a Annual diffusive CH4 emissions, averaged across an ensemble 100,000 individual calculations using the artificial neural network mapping method. b Same as a, but using the random regression forest mapping method
Fig. 4
Fig. 4
Regional and global diffusive CH4 emissions. a Violin plot for annual diffusive CH4 emissions integrated across four bathymetric regions, computed using Monte Carlo method to propagate uncertainty in ∆CH4 and gas transfer velocity. Violin thickness corresponds to probability density, with think black lines at 25th and 75th percentiles, thick line at median, and diamond at mean value. Light gray shading for each region spans the 10–90th percentiles for all estimates, combining artificial network (ANN) and random regression forest (RRF) ensembles. b Probability density functions for globally integrated CH4 emissions from ANN and RRF methods. Diamonds and light gray shading as defined in a
Fig. 5
Fig. 5
Ebullitive and total CH4 emissions. a Modeled transfer efficiency of CH4 in bubbles from the seafloor to surface ocean, for 2 and 8 mm diameter bubbles, and integrated across a characteristic bubble size spectrum (Supplementary Fig. 7). Diamond and circle points represent the mean transfer efficiency for bubbles released uniformly between 0–100 and 0–200 m, respectively, and gray shading marks the range of 11–17% bounded by these cases. b Probability density functions for total oceanic CH4 emissions, combining the distribution for diffusive fluxes (Fig. 4b) with two uniform probability distributions for ebullitive emissions that are obtained by applying 11–17% transfer efficiency to seafloor ebullition rates of 35 and 18–48 Tg yr−1. Dark and light gray shading mark the likely range (10–90th percentiles) for the two estimates
Fig. 6
Fig. 6
Controls on surface ocean ∆CH4. a Joint probability distribution for mapped ∆CH4 and seafloor depth (zsf) in coastal ocean regions (<2000 m depth). Color scale represents the frequency of gridcells with a given combination of log10(depth) and log10(∆CH4), after averaging together all 200,000 machine-learning maps. Black line is the best fit for the mapped data (∆CH4 = 67zsf−0.7, R2 = 0.55). b Scatter plot of observed ∆CH4 versus depth. Gray points show raw data; black circles with errorbars show mean ± s.d. ∆CH4 within depth bins. Red line is best fit to the binned data (CH4 = 69zsf−0.8, R2 =0.94). c, d Same as a and b, but for the relationship of mapped c and observed d ∆CH4 to net primary production (NPP) in open ocean (>2000 m depth) environments. In c, black line is the best fit for mapped data (∆CH4 = (0.5NPP − 62)/103, R2 = 0.30), and symbols represent large-scale averages (Supplementary Fig. S8). In d, black circles show mean ± s.d. ∆CH4 within NPP bins, and red line is best fit to the binned data (∆CH4 = (0.3NPP + 14)/103, R2 = 0.91)
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Hartmann, D. L., Klein Tank, A. M. G., Rusticucci, M. & Alexander, L. V. in Climate Change 2013 – The Physical Science Basis: Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (ed. Change Intergovernmental Panel on Climate) 159–254 (Cambridge University Press, 2014).
    1. Etminan M, Myhre G, Highwood EJ, Shine KP. Radiative forcing of carbon dioxide, methane, and nitrous oxide: a significant revision of the methane radiative forcing. Geophys. Res. Lett. 2016;43:12614–12623. doi: 10.1002/2016GL071930. - DOI
    1. Hamdan LJ, Wickland KP. Methane emissions from oceans, coasts, and freshwater habitats: new perspectives and feedbacks on climate. Limnol. Oceanogr. 2016;61:S3–S12. doi: 10.1002/lno.10449. - DOI
    1. Saunois, M. et al. The global methane budget: 2000–2012. Earth Syst. Sci. Data Discuss. 1–79, 10.5194/essd-2016-25 (2016).
    1. Barnes RO, Goldberg ED. Methane production and consumption in anoxic marine sediments. Geology. 1976;4:297–300. doi: 10.1130/0091-7613(1976)4<297:MPACIA>2.0.CO;2. - DOI

Publication types