Isotopic constraints on methanes global sources and ENSOdependence

  • Slides: 16
Download presentation
Isotopic constraints on methane’s global sources and ENSO-dependence Hinrich Schaefer S. Mikaloff Fletcher, G.

Isotopic constraints on methane’s global sources and ENSO-dependence Hinrich Schaefer S. Mikaloff Fletcher, G. Brailsford, T. Bromley, R. Martin; Sylvia Nichol; NIWA C. Veidt, I. Levin; University of Heidelberg B. Vaughn, S. Englund, J. Miller, J. White; INSTAAR E. Dlugokencky, NOAA-ESRL K. Lassey, D. Lowe; SGS* *Society of Gentlemen Scientists

Isotopic constraints on methane’s global sources and ENSO-dependence Schaefer et al. , Science, 2016

Isotopic constraints on methane’s global sources and ENSO-dependence Schaefer et al. , Science, 2016

Basic study design: δ 13 CH 4 record • Create globally averaged annual δ

Basic study design: δ 13 CH 4 record • Create globally averaged annual δ 13 CH 4 from records from various laboratories

Basic study design: box model • Small changes in δ 13 CH 4 of

Basic study design: box model • Small changes in δ 13 CH 4 of global source • Emissions decreased after 1993; increased after 2006 • Use 1 -box model to identify δ 13 CH 4 of “lost” emissions causing plateau and of “new” emissions for renewed rise • Identify isotopic leverage of “changed” emissions (~ 20 Tg/a) of total (~550 Tg/a) emissions

Basic study design: box model • Small changes in δ 13 CH 4 of

Basic study design: box model • Small changes in δ 13 CH 4 of global source • Emissions decreased after 1993; increased after 2006 • Use 1 -box model to identify δ 13 CH 4 of “lost” emissions causing plateau and of “new” emissions for renewed rise • Identify isotopic leverage of “changed” emissions (~ 20 Tg/a) of total (~550 Tg/a) emissions

Basic study design: box model Emission rate Source 13 C [CH 4] Atmosphere 13

Basic study design: box model Emission rate Source 13 C [CH 4] Atmosphere 13 C Removal rate Sink Calculates atmospheric 13 C from: • a global source and its 13 C/12 C • Sinks and 13 C-fractionation Fractionation Modelling approach: • For given sink, determine source history by inverse run • Pick onset of an event • In forward run, force continuation of previous trend (“base run”) • overlay source (or sink) “perturbation” to match atmospheric history Turn a faucet with set temperature on or off.

Findings I Plateau caused by “lost” emissions with δ 13 CH 4 of ~

Findings I Plateau caused by “lost” emissions with δ 13 CH 4 of ~ -40‰ ► Less fossil-fuel CH 4 (collapse of Soviet gas production? )

Findings I Plateau caused by “lost” emissions with δ 13 CH 4 of ~

Findings I Plateau caused by “lost” emissions with δ 13 CH 4 of ~ -40‰ Alternatively, OH-variability (Montzka et al. , 2011) Montzka et al. , 2011 explains both [CH 4] and δ 13 CH 4

Findings II • “New” emissions with δ 13 CH 4 of ~ -60‰ cause

Findings II • “New” emissions with δ 13 CH 4 of ~ -60‰ cause renewed rise • No (minor) fossilfuel CH 4 • Biogenic source • Wetlands • Agriculture

Findings II • “New” emissions with δ 13 CH 4 of ~ -60‰ cause

Findings II • “New” emissions with δ 13 CH 4 of ~ -60‰ cause renewed rise • Prescribed OHtrend (-0. 15%/yr) matches [CH 4], but not δ 13 CH 4

Advantages/limitations of study • No seasonal information (Martin Manning’s talk) • No geographic (latitudinal)

Advantages/limitations of study • No seasonal information (Martin Manning’s talk) • No geographic (latitudinal) information (inverse models) But: • Sensitive to small changes in total source signature • Independent of prior estimates • Emission stacks (e. g. Rice et al. , 2016) • Isotopic signature of sources (Schwietzke et al. , 2016) Finally: • Subject to same ambiguities from under-constrained system as other studies

Points regarding newer findings • Role of OH-variability (sink) • Confirmed by Mc. Norton

Points regarding newer findings • Role of OH-variability (sink) • Confirmed by Mc. Norton et al. (2016) for plateau onset • Controversial for renewed rise (Turner et al. , 2017; Rigby et al. , 2017) • Role of fossil-fuel methane • Increase during plateau postulated by Rice et al. , 2016 • Ethane trends: consistent with decrease during plateau (Simpson et al. , 2012); but increase since 2010 (Helmig et al. , 2016; and other studies) • Nature of biogenic source • Agriculture • (Tropical) wetlands (Nisbet et al. , 2017)

ENSO correlation with δ 13 CH 4 El Nino Southern Oscillation (ENSO): • Controls

ENSO correlation with δ 13 CH 4 El Nino Southern Oscillation (ENSO): • Controls weather in regions of tropical CH 4 production • Opposite influence on wetlands and biomass burning • Reinforcing influence on δ 13 CH 4 • La Nina dominant after 2007 El Nino (cool, dry): Biomass burning CH 4 (~-20‰) Wetland CH 4 (~-58‰) La Nina (warm, wet): Biomass burning CH 4 Wetland CH 4

ENSO correlation with δ 13 CH 4 and HCN • HCN: biomass burning proxy

ENSO correlation with δ 13 CH 4 and HCN • HCN: biomass burning proxy • ENSO indices: • SOI: sea level pressure • ONI: sea surface temperature • MEI: multi-variate HCN 12 -24 gro vs ONI run 150. 0 R 2 = 0. 398 100. 0 50. 0 -1. 50 -1. 00 -0. 50 Latitudes: 0. 00 -50. 0 0. 50 1. 00 -100. 0 -150. 0 • S. Tropics (Ascension Island, Samoa) • S. mid-latitudes (Baring Head, NZ) • Global SOI run vs BHD gro 0. 150 R 2 = 0. 1625 0. 100 0. 050 Allowing for different lag times -3. 00 -2. 00 -1. 00 0. 00 -0. 050 -0. 100 1. 00 2. 00 3. 00 4. 00

Rate of δ 13 CH 4 variability explained by ENSO: • S. tropics: 0%

Rate of δ 13 CH 4 variability explained by ENSO: • S. tropics: 0% - 32% (SOI) • S. mid-latitudes: 5% - 22% (SOI) For HCN: up to 30% - 40% of growth rate (MEI, ONI) • Global: 2% - 20% (SOI) ► ENSO is minor driver of CH 4 cycle, ► Stronger control on biomass burning than wetlands SOI (running mean) ONI (running mean) Lag r 2 (months) 0 -100 km growth rate 0. 24 2 0. 32 2 12 -24 km growth rate 0. 26 1 0. 40 1 global growth rate 0. 11 2 0. 07 10 BHD growth rate 0. 16 0 0. 05 0 ASC growth rate 0. 23 0 0. 08 1 ASC running mean 0. 32 26 0. 09 6 SMO running mean 0. 22 20 0. 08 17 HCN d 13 CH 4

Parting thought Emission rates (for stable OH): • Before 1992; after 2007: • low

Parting thought Emission rates (for stable OH): • Before 1992; after 2007: • low inter-annual variability (IAV). • Same trend? • 1993 – 2006: high IAV; generally lower With OH-variability (1994 – 2007; Montzka et al. , 2011): • Some “improved” IAV during plateau • Some “worse” IAV afterwards ► what causes the break in IAV pattern? ► did it cause the plateau?