A revised look at the oceanic sink for atmospheric carbon tetrachloride (CCl 4 )

Transcription

1 A revised look at the oceanic sink for atmospheric carbon tetrachloride (CCl 4 ) James H. Butler 1, Shari A. Yvon-Lewis 2,6, Jürgen M. Lobert 3,6, Daniel B. King 4,6, Stephen A. Montzka 1, James W. Elkins 1, Bradley D. Hall 1, Valentin Koropalov 5, John L. Bullister 7, Lei Hu 7, Yina Liu SPARC Carbon Tetrachloride Workshop 5,6 October NOAA Earth System Research Laboratory, Global Monitoring Division, Boulder, CO, USA 2. Texas A&M University, Department of Oceanography, College Station, TX, USA 3. Entegris Inc., Franklin, MA, USA 4. Chemistry Department, Drexel University, Philadelphia, PA, USA 5. Roshydromet, Moscow, Russia 6. CIRES, University of Colorado, Boulder, CO, USA 7. NOAA Pacific Marine and Environmental Laboratory, Seattle, WA, USA 8. Pacific Northwest National Laboratory, Richland, WA 99354

2 Why is this important? CCl 4 is a strong ozonedepleting gas for which most production has ceased. Its rate of decline is slower than its limited production and atmospheric lifetime suggest. The oceanic sink is typically treated as a significant contributor to the lifetime of CCl 4 in the atmosphere, along with reaction in the stratosphere and loss to soils.

3 Purpose of this study Re-examine the oceanic sink to provide more confidence in our ability to estimate the rate of atmospheric CCl 4 removal by the ocean. With data from 17 cruises, this allows us to provide a much more representative picture of oceanic removal rates. o 17 cruises o (Almost) All oceans o All seasons o 23 years ( )

4 Sampling Ø Surface Air, Surface Equilibrator (Weiss) Ø Hydrocast samples at surface o Some taken at depth Analysis Ø ECGC, GCMS Ø Direct, Purge and Trap How did we do it?

5 What did we find out? CCl 4 is undersaturated in the surface ocean nearly everywhere, virtually all the time. This undersaturation exceeds that which might be expected from physical effects, such as mixing of water masses. The undersaturations are largely similar Ø Exceptions appear to be areas of upwelling Southern Ocean (BLAST 3; 1996) East Pacific (BLAST 1; 1994) Atlantic, Coastal Pacific (GasEx-98; 1998) Equatorial (SAGA-3; 1990) East Atlantic (HalocAST-A; 2010) East & West Atlantic (BLAST )

6 Are we sure? East Pacific HaloCAST-P; 2010) Surface samples from hydrocasts (circles) vs. equilibrator measurements (spikes) suggest no sampling bias Often, but not always, influences of physical effects make the anomaly positive or less negative. CFC-11 is used to correct them, Observed Anomaly (PHASE; 2004) Corrected Anomaly (PHASE; 2004)

7 Correction for physical effects Warming or cooling of surface waters Mixing of water masses Air injection All are dependent on solubility and diffusivity of gas Ø Thus, temperature and salinity Δ CFC-11 SAGA II, 1987

8 How we calculate the oceanic Fg= KwgA/Hg ( pgw pga/pga ) sink p a F g, Δg(%)=100( pgw pga/pga ) p w FCT,corr= KwgApCT,a/HCT ( ΔCT Δf11/100 ) τo,ct= Ma,CT/FCT corr,o = ntrct/ rctfct,o Fg=net flux across ocn surface Kwg=f(u,T) A =area Hg=f(S,T) p=partial pressure τ o =lifetime wrt oceanic loss M a =Moles of CT in atmosphere n tr =Moles of CT in troposphere r =fraction CT in troposphere

9 Uncertainty and Confidence Uncertainty Ø Air-Sea Exchange Coefficient, K w (u,t) Ø Saturation Anomaly, Δ(%) Ø Solubility Confidence Ø Multiple K w relationships Ø Distribution of Saturation Anomaly Ø Physical corrections Ø Distribution of Flux

10 Histogram of Δ(%)

11 Averages of all Cruises Median = -6.3% Mean = -7.1%

12 Skinning the Cat... Bin mean median N st dev std error < to to to to to to to to to to to to > Mean (bins) = -6.6 ± 1.0 Median (bins)= -5.8 ± 0.8 Mean (all pts) = -7.1 ± 0.2 Median (all pts) = -6.3

13 Binned Δ(%) s Mean (bins) = -6.6 ± 1.0 Median (bins)= -5.8 ± 0.8

14 Air-Sea Exchange c.v. = 29%

15 Air Sea Exchange Coefficients Relationship Kw(CCl4) m/d Liss and Merlivat (1986) 1.6 Wanninkhof (1992) 3.4 Nightingale (2000) 2.2 Sweeney (2007) 2.3 Wanninkhof (2009) 2.0 Mean 2.3 SD 0.7

16 What s causing this undersaturation? 35.5ºS 40.5ºN Relative concentrations of CCl 4 are consistently less than CFC-11 at intermediate depths, suggesting consumption as oxygen declines.

17 Subsurface Activity CCl 4 consumed completely at near-zero oxygen CFC-11 much less so (CFC-12 not at all)

18 Sinks and Lifetimes Scenario F(ocn) Fn Fs τ(ocn) *τ(atm) Fs/Fn *Loss to Ocn Gg/y Gg/y Gg/y yr yr % Global MEDIAN corrected SA Global MEAN corrected SA Latitude binned MEDIAN corrected SA Latitude binned MEAN corrected SA MEAN SD *τ(atm) and Percent loss to Ocean assume τ(strat) = 44 y, and τ(soils) = 245 y (from board yesterday)

19 What did we find out? The oceanic sink is responsible for removing ~14% of the CCl 4 from the atmosphere, representing a partial atmospheric lifetime of 236( )y Considering this sink and the removal of CCl 4 in the stratosphere, the mid-range estimate of the atmospheric lifetime of CCl 4 would be 32±1y (formerly 26y in the WMO/UNEP Scientific Assessments). (Note: uncertainties are for the ocean sink only)

20 What does this all mean? Irreversible removal of CCl 4 by processes within the ocean has a significant impact (~14%) on the lifetime of CCl 4 in the atmosphere CCl 4 removal could take place in the surface ocean, but there is considerable evidence in depth profiles that it is removed more rapidly at depth near the oxygen minimum. South Pole The influence of the oceanic sink on the atmospheric lifetime is robust though smaller than previously suggested. The current best estimate of the atmospheric lifetime of CCl 4 is 32 years All Stations

21

22 Backups

23 Assume first-order loss drives the sink Fg,degr= kminpgaz/hg ( 100+Δg/100 ) p a F g,as F g,degr dc/dt = kminc= kminpga/hg ( 100+Δg/100 ) p w kmin= Kw/z (Δf11 Δg)/(100+Δg) z=mixed layer depth

24 Physical properties affecting airsea exchange CFC-11 CFC-12 CCl4 Physical Properties Diffusivity (D; 10 5 cm 2 s -1 ) Solubility (H; m 3 atm mol -1 ) ΔH/ΔT (0-30 C)