12.842 Climate Physics and Chemistry Fall 2006 …dspace.mit.edu/.../0/part2_2pdf.pdfClimate Physics and Chemistry Fall 2006 Natural and Fossil Fuel CO 2 in the Ocean (II) Direct measurements

12.842Climate Physics and Chemistry

Fall 2006

Natural and Fossil Fuel CO2 in the Ocean (II)

Direct measurements of CO2 in the atmospherefor the past 47 years

Image courtesy of DOE.

Atmospheric CO2 growth rate

01955

Going up. Annual global rates of CO2 accumulation in the atmosphere and CO2 emission from fossil fuel burning since 1958.

1965 1975 1985

2

4

Pg C

arbo

n/ye

ar 6

Annual emissions from fossil fuels

Linear average CO2 growth rate


8

1995 2005

Figure by MIT OCW based on Science 298: 2344.

Fossil fuel CO2 absorption by the ocean: how much, how fast?

How fast does the ocean absorb CO2? What are the important rate- and capacity-limiting processes? How might these processes change in the future?

• Direct observation of changing ocean CO2 concentrations (“the oceanic Keeling Curve”): this method is difficult because the ocean response is 10% of the atmospheric forcing. The CO2content of the atmosphere has increased by 20% over the past 45 years, so ocean water in equilibrium with the atmosphere has only increased by 2%. And most of the absorbed CO2has been mixed into a large pre-existing background in the deep ocean, so the signal is much less than 2%.

• Estimation by subtracting out the natural background, based on “Redfield Ratio”stochiometry. This method relies on the higher accuracy of recent measurements and the stoichiometry of the oxidation of marine organic matter. Works reasonably well for North Atlantic, more difficult for Southern Ocean and Pacific (higher background, less certain initial conditions).

• Tracer method, combined with simple robust models: tracers such as bomb tritium and chlorofluorocarbons enter the ocean from the atmospere, much as does CO2. We can safely assume that water that has tritium and CFCs also has anthropogenic CO2. The method requires some correction for the chemical differences between these tracers and for the different time histories of emissions. Nonetheless, these corrections are straightforward and the model-dependence is not strong.

Karl ALOHA pCO2 time-series

Courtesy of David Karl, University of Hawaii at Manoa. Used with permission.

The Revelle Factor

Ocean uptake of fossil fuel CO2:

two limiting cases and the real situation

Fundamental factors controlling ocean fossil fuel CO2 uptake

• Physical transfers:– CO2 crosses the air-sea interface

- Chemical equilibrium determines total possible transfer(carbonic acid equilibrium constants, summarized by Revelle Factor)

- Gas exchange dynamics across the air-sea interface determines the rate of approach to chemical equilibrium. Gas exchange is a functionof wind speed and other factors.

- Estimated from bomb 14C uptake and 222Rn deficit

– CO2 that dissolves into the surface mixed layer is carried by ocean circulation into the upper ocean and deep sea.

• How fast does this occur, and what factors affect the rate of circulation?

Outline:• Estimating gas exchange rates

• Thin film model• Piston velocity• Wind speed dependence

• Ocean CO2 exchange• Global CO2 exchange rate• Air-sea disequilibrium and spatial variability

• Estimating the penetration of ocean surface waters into the deep ocean• Ocean circulation primer• Transient tracers

• Estimating ocean uptake of fossil fuel CO2: is there a “missing sink”?• The ΔO2/ΔCO2 method for estimating ocean and biospheric fossil fuel

uptake: quantifying the “missing sink”• Other methods for observing fossil fuel CO2 uptake

• δ13C• “excess CO2”• Historical proxy methods: δ13C in tree rings and corals

• Extrapolating fossil fuel into the future





• Estimating ocean uptake of fossil fuel CO2: is there a “missing sink”?• The ΔO2/ΔCO2 method for estimating ocean and biospheric fossil fuel uptake:

quantifying the “missing sink”• Other methods for observing fossil fuel CO2 uptake



Stagnant Film Model

of Gas Exchange

Image removed due to copyright restrictions. Citation: Figure 3-2. Broecker, W. S., and T. H. Peng. Tracers in the Sea.Eldigio Press Lamont Doherty Geological Observatory, 1982, 690 pages.

Note: diffusion is random in direction and gas exchange occurs in both directions. A net flux occurs when there is a concentration gradient where there are more atoms going in

one direction than in the other.

Piston velocity• D/z has dimensions of velocity• Gas exchange diffusion through the stagnant film acts as if two

pistons, moving in opposite directions, are sweeping the dissolved gases in and out of the stagnant film.

• Piston velocity for CO2 in the ocean is about 2000 m/yr

stagnant film

diffusion from water into air

diffusion from airinto water

222Rn deficit method for determining

gas exchange rates

Image removed due to copyright restrictions. Citation: Figure 3-3. Broecker, W. S., and T. H. Peng.Tracers in the Sea. Eldigio Press Lamont Doherty GeologicalObservatory, 1982, 690 pages.

Global average CO2 exchange rate from natural radiocarbon

• 14C is formed at the top of the atmosphere by the interactions of cosmic rays with atmospheric gases

• This 14C is then (well) mixed in the atmosphere and enters the ocean.

• The decay rate of 14C in the ocean + atmosphere + biosphere ≈ ocean

is equal to the production rate.• Hence the flux of 14C into the surface ocean must be equal

to the difference in radiocarbon concentration between the atmosphere and mixed layer of the ocean.

• So we can calculate zfilm from natural radiocarbon.

Gas exchange increases with increasing wind speed

18151296300 0

225

150

75

CFT: A

DTM: BDTM: C

14C: D14C: E

222Rn: F222Rn: G222Rn: H

222Rn: I

20

40

60

80

Plot of kL versus U10 for representative oceanic measurements of kL made using the controlled flux technique (CFT), the deliberate tracer method (DTM), bomb 14C invasion rates, and the 222Rn method. All measurements have been normalized to Sc=600 assuming that kL The wind speed parameterizations of Wanninkhof [1992] (solid line) and Liss and Merlivat [1986] (dashed line) are also shown. Data keyshown on the figure, Where A is the CFT data of Haubecker ans Jahne [1995], B is the DTM data of Wanninkhof et al. [1993], C is the DTM data of Watson et al. [1991], D is the bomb 14C data from Cember [1989], E is the bomb 14C data from Broecker et al. [1985], F is the 222Rn data from Kromer and Roether [1983], G is the 222Rn data from Peng et al. [1979], H is the 222Rn data from Peng et al. [1974], and I is the222Rn data from Broecker and Peng [1971].

Sc-1/2

k L (x

10-6

m s-1

) (S

c=60

0)

kL (cm b

-1) (Sc=600)

Figure by MIT OCW.

Surfactants can also affect gas exchange

Image removed due to copyright restrictions. Citation: Figure 4. Asher and Wanninkhof. JGR 103 (1998): 15939.

Recent estimates of gas exchange rateas a function of wind speed

Image removed due to copyright restrictions. Citation: Figure 8. McGillis, W. R., et al. Marine Chemistry 75 (2001): 267-280.









Rate of exchange of CO2 between ocean and atmosphere:

2000 m yr-1 * 10-5 moles kg-1 * 1000 kg m-3 = 20 moles m-2 yr-1

Piston

velocity

conc. ofgaseous

dissolved CO2

conversionfactor

Exchange rate of CO2across air/sea interface

CO2 can show strong short-scale variability (e.g. the North Atlantic ocean)

Image removed due to copyright restrictions. Citation: Figure 1. Watson, Andrew J., C. Robinson, J. E. Robinson, P. J. Leb. Williams, andM. J. R. Fasham. “Spatial Variability in the Sink for Atmospheric Carbon Dioxide in the NorthAtlantic 50.” Nature 350 (March 7, 1991).

Takahashi CO2 flux estimate

Image removed due to copyright restrictions. Citation: see image about mean annual flux for 1995 on http://www.ldeo.columbia.edu/res/pi/CO2/.

Estimating Ocean of Fossil Fuel CO2 Uptake

• CO2 in the atmosphere equilibrates on a timescale of less than a year. Hence this is not a rate-limiting step (mean life concept).

• The rate-limiting step is the penetration of surface waters into the interior of the ocean.

• In order to estimate ocean CO2 uptake, we need to find out how fast surface waters move into the deep sea.

The carbon “solubility pump”

The solubility pump: CO2 is more soluble in cold waters than in warm waters (the thermal pump), and more soluble in fresher waters than saltier waters (the salt pump). If alkalinity were uniform throughout the ocean and if both cold and warm surface waters equilibrated their pCO2 with the atmosphere, then cold surface waters would have a higher dissolved carbon dioxide content than warm surface waters. As these cold surface waters circulate into the deep interior of the ocean, deep waters will have more CO2 than warm surface waters.

Gas solubility

Source: Sarmiento and Gruber (2006)

CCl4

CO2

CFC-13

N2O RnCFC-12

100,000

80,000

60,000

40,000

20,000Solu

bilit

y (m

mol

m-3

atm

-1)

00 5 10 15 20 25 30

Kr

ArCH4

O2

SF6

N2Ne HeSo

lubi

lity

(mm

ol m

-3 a

tm-1

)

00 5 10 15 20 25 30

4,000

3,000

2,000

1,000

Plot of the solubility of various gases as a function of temperature. The solubility of all gases decreases with increasing temperature, but there exist large differences in the solubility of the different gases for a particular temperature. These differences can be understoodin terms of their molecular weight (ideal gas) or other factors, such as molecular interactions between the gas and the water (nonideal gas).Note the change in vertical scale between panels (a) and (b). Based on the empirical functions listed in table 3.2.2.

Temperature (oC)Temperature (oC)(a) (b)

Figure by MIT OCW.

The carbon “biological pump”• The biological pump: organisms remove carbon and nutrient elements from the surface ocean (which is equilibrated with

atmospheric oxygen; note oxygen solubility is a function of temperature); the debris from these organisms sinks and decomposes, releasing carbon and nutrient elements into the deep water and consuming oxygen.

– classical Redfield Ratio:

• (CH2O)106(NH3)16(H3PO4) + 138 O2 -> 106CO2 + 16HNO3 +H3PO4

• This stoichiometry proposed by Redfield is based on elemental ratios he observed in ocean water samples and plankton. It considers marine organic matter as if it were a mixture of carbohydrates (CH2O), proteins (containing NH3), and phospholipids and nucleic acids (H3PO4 bearing). In reality, a broad mixture of compounds occur, and the observed stoichiometry ΔO2:ΔC in deep ocean waters implies a higher value (~165), because of the presence more hydrocarbon-like functional groups:

• (CH2O)111(CH4)11(NH3)16(H3PO4) + 165 O2 = 122 CO2 + 16 HNO3 + 149 H2O + H3PO4

• Note the production of nitric and phosphoric acid in this process; this acid changes alkalinity. Once you acknowledge this process, you also need to take into account another effect that reduces the acid effect on alkalinity by about 1/3: the presence of ion-exchanged carboxyl groups:

– (CH2O)102(CH4)14(HCOO-Na+)6(NH3)16(H3PO4) + 165 O2

» = 122 CO2 + 16 HNO3 + 146 H2O + H3PO4 + 6 NaOH

• There is debate on whether the C:P and N:P stoichiometries are fundamental to marine ecosystems, or whether there is some plasticity (e.g., could N:P = 25?). (There is less debate on variable C:N ratios than for C:P).

• We should also note that some organisms precipitate inorganic shells out of calcium carbonate (CaCO3) and silica (SiO2.nH2O).

For example for every 106 organic carbon atoms converted to organic matter, about 22 CaCO3 molecules are precipitated (ocean-wide average).









Practical implications of ocean vertical thermal gradient

• This experiment, which seem’d first but mere food for curiosity, became in the interim very useful to us. By its means we supplied our cold bath, and cooled our wines or water at our pleasure; which is vastly agreeable to us in this burning climate.

Ellis (1751)

Western North Atlantic Potential Temperature Section

Image courtesy of U.S. Government.

Transient Tracers• Other substances have been added to the atmosphere and surface ocean in

the recent past: bomb tritium (3H) and 14C, and chlorofluorocarbons (CFCs). The difference between these tracers and CO2 is that they begin entering into an ocean where there did not previously exist (or at very low concentrations). We can then use these substances to estimate how much of the ocean has been in contact with the surface in recent years.

Tritium in the North Atlantic Ocean

Image removed due to copyright restrictions. Citation: Ostlun, H. G., and C. G. H. Rooth. “The North Atlantic Tritium and Radiocarbon Transients 1972-1983.” J Geophys Res 95 (1990): 20147-20165.

CFC 11 in the North Atlantic Ocean

Image removed due to copyright restrictions. Citation: Image of Atlantic CFC-11 section. Doney, S. C., and J. L. Bullister.“A Chlorofluorocarbon Section in the Eastern North Atlantic.” Deep-Sea Res 39 (1992): 1857-1883.

Ocean uptake of fossil fuel CO2:

two limiting cases and the real situation









Calculating the ocean fossil fuel CO2 uptake: Step 1: subtract out the steady state natural cycle

01955

Going up. Annual global rates of CO2 accumulation in the atmosphere and CO2 emission from fossil fuel burning since 1958.

1965 1975 1985

2

4

Pg C

arbo

n/ye

ar 6

Annual emissions from fossil fuels

Linear average CO2 growth rate


8

1995 2005

Figure by MIT OCW based on Science 298: 2344.

Calculating the ocean fossil fuel CO2 uptake: Step 2: calculate the mean age of fossil fuel CO2

• The concept of mean age works as if we were to attach a clock to every atom of CO2 emitted (set to zero at the time of emission), then at some later point gather in all the clocks and calculate the average.

Mean Age =−tF(t)dt− t

0∫F(t)dt− t

0∫

where t = 0 is the presentand F(t) = CO2 emissions

The mean age of CO2 at present is about 28 years

IPCC CO2 budget, 1980-89

Table 1. IPCC global CO, budget for 1980-1989 [as adapted from (38) by (56)].

Reservoir Average flux)(Gt of C year

Sources

Sinks

Imbalance (sources - sinks)

Fossil fuelsDeforestation and land use

Total

AtmosphereOceans (modeled uptake)

Total

5.41.6

0.51.0

+_+_

7.0 1.2

3.22.0

0.10.8

5.2 0.8

1.8 1.4

+_+_

+_

+_

+_

-1

Other ways to estimate ocean CO2 uptake

• CO2 influx/efflux estimate (from ΔpCO2)

• Loss of oxygen from the atmosphere

• δ13C of oceanic CO2 (because fossil fuels are ~ -27‰ compared to ~0‰ for the ocean) O2 influx/efflux estimate (from ΔpCO2)

• “excess CO2” calculation based on ocean CO2 and nutrient data





• Estimating ocean uptake of fossil fuel CO2: is there a “missing sink”?• The ΔO2/ΔCO2 method for estimating ocean and biospheric fossil fuel

uptake: quantifying the “missing sink”• Other methods for observing fossil fuel CO2 uptake



Atmospheric O2 decrease, 1991-1995

Source: Keeling et al., 1996.

91-240

δ(O

2/N

2) (p

er m

eg)

δ(O2/N2) of air samples collected at Alert, La Jolla and Cape Grim and measured using the interferometric method.

-200

-160

-120

-80

-40

0

92Year93 94 95

Alert (82.5oN, 62.3oW)Cape Grim (40.7oS, 144.7oE)La Jolla (32.9oN, 117.3oW)

Figure by MIT OCW based on Nature Vol. 381, 1996.

Atmospheric ΔCO2-ΔO2 vector diagram

Source: R. Keeling lab

-190350

(CO2)GLOB (p.p.m.)

δ(O

2/N

2)G

LOB

(per

meg

)

352

Atmosphericincrease

Land BioticUptake

Ocean Uptake

354 356 358 360 362 364 366 368 370

-170

-150

Jul 91Jan 92Jul 92Jan 93Jul 93

Jan 94

Jul 94

Change calculated from fossil fuel burning alone-130

-110

-90

Ocean uptake

Land biotic uptake

Figure by MIT OCW based on Keeling, et al., 1996.

Another O2-CO2time series

Image removed due to copyright restrictions. Citation: Figure 1. Battle, Bender, and Tans, et al. Science 287 (March 31, 2000): 2467.

Using O2 and carbon

isotopes to

estimate fossil

fuel CO2uptake

Image removed due to copyright restrictions.Citation: Figure 4. Battle, Bender, and Tans, et al. Science 287 (March 31, 2000): 2467.

Increasing amplitude and length of the growing season

Source: Keeling and Chin (1996) Nature 382, 146-149.

-10

-8

-6

-4

-2

0

2455 60 65 70 75 80 85 90 95

Day

s (re

lativ

e to

day

212

)

55

1.351.301.251.201.15

1.101.051.000.95

60 65 70 75 80 85 90 95

Rel

ativ

e am

plitu

de

Year YearTrends in relative amplitude and timing of the seasonal cycle of atmospheric CO2. a, At Mauna Loa Observatory, Hawaii. Annual values of the amplitude (Left plot, dots connected by orange lines) were determined from weekly averaged continuous concentration data15 fitted annually to a phase-locked 4-harmonic seasonal function. This function together with a linearly increasing gain factor, was first established by a fit to the full record3, after which the gain was redetermined separately for each year. The linearly increasing gain factor, referenced to 1964, was computed to be 0.675 + 0.045% per year. A smoothing spline19 (Purple curve, with standard error, σA, of 0.028%) shows quasi-decadal changes in the relative amplitude. The timing of the downward zero crossing of the seasonal cycle (Right plot, dots connected by solid line segments), was obtained from 3-year fits to the CO2 record.

_

Figure by MIT OCW.









Atmospheric δ13C of CO2, 1976-1995

Image removed due to copyright restrictions. Citation: Figure 13. Keeling, et al. Glob Biogeochem Cycles 10 (1996): 335.

Temperature-dependent carbon

isotope equilibration with the ocean: now and pre-anthropogenic

Changing δ13C of the ocean due to fossil fuel uptake

Image removed due to copyright restrictions. Citation: Figure 1 and Figure 2. Quay, P. D., B. Tilbrook, and C. S. Wong.Science 256 (April 3, 1992): 74-79.

Evolving Pacific Ocean δ13C depth profiles

Image removed due to copyright restrictions. Citation: Figure 3. Quay, P. D., B. Tilbrook, and C. S. WongScience 256 (April 3, 1992): 74-79.

Using ice core CO2 and tree-ring δ13C data to estimate anthropogenic carbon emissions

Image removed due to copyright restrictions. Citation: Figure 10-2. Broecker, W. S., and T. H. Peng. Tracers in the Sea.Eldigio Press Lamont Doherty Geological Observatory, 1982, 690 pages.

“Excess CO2” method:(over) simplified concept

P

TCO2

natural CO2-P (Redfield Ratio)

excess anthro pogenic CO2

“Anthropogenic” CO2 E-W section in the North Atlantic









Modeling the future evolution of anthropogenic

CO2Image removed due to copyright restrictions. Citation: Figure 10-14. Broecker, W. S., and T. H. Peng.Tracers in the Sea. Eldigio Press Lamont Doherty Geological Observatory, 1982, 690 pages.

CO2 projections for the next 1000 years

Image removed due to copyright restrictions. Citation: Figure 1. Caldiera and Kasting. Nature 366 (1993): 251-253.

Projection of CO2 for the next 40,000 years

source: Archer et al. (1996)

0

A23

A22

A22

A

B

2000 4000 6000 8000 10k 40k

500

1000

Atm

osph

eric

pC

O2

ppm

v

1500

2000

∞

CaCO3no CaCO3

Figure by MIT OCW. Adapted from Archer et al., 1997. Similar scenarios have beenpresented previously by Eric Sundquist.

Ocean acidification:another consequence of increasing CO2

• As CO2 increased from 280 to 375 ppmV in the atmosphere, tropical ocean surface water pH dropped by 0.10 pH unit. When CO2 is doubled, pH will drop another 0.14 unit.

• This drop in pH decreases the supersaturation of the upper ocean with respect to CaCO3.

• The rate of calcification by corals and other carbonate-secreting organisms decreases as saturation decreases.

Alternatives to continued fossil fuel CO2 emissions• Conservation

– One could envision perhaps a factor of two improvement in fuel consumption (hybrid vehicles) and continuing improvements in other energy uses. But it would be extremely difficult for conservation alone to solve the problem.

• Find other energy sources– Most alternatives have limited potential at the present; future developments may improve

their viability:– Wind Power: now economically viable to some extent given government subsidy.

Problems: NIMBY, requires backup when winds are weak.– Solar: Not economically viable now except in remote areas or given government subsidy

(California). There is potential for efficiency improvement, but as for wind power, requires backup or storage for night and when overcast.

– Nuclear: This option is probably the only one that could reliably provide large scale base power. It cannot directly serve transportation, although it could be enlisted as a source of hydrogen or to charge (hypothetical) improved batteries for electric cars. In some regions it is important now (e.g. France). But in other regions safety concerns have made it uneconomic (e.g. U.S.). The safety issue possibly may be overcome, but the large scale introduction of nuclear power would worsen its own environmental problem: what do we do with spent fuel that remains radioactive for tens of thousands of years? There is no demonstrated solution yet, although demonstration projects (e.g. Yucca Mountain) are underway.

Alternatives #2• Carbon sequestration

The idea is to capture CO2 (most efficiently done at the site of fossil fuel consumption) and then place it somewhere where it will not reach the atmosphere for thousands of years.

• Capture is done most efficiently at large power stations; it is difficult to impossible for small or mobile sources.

• Capture can be achieved by several technologies: quaternary amines can retain carbon and the release it; power plants could be fueled by 100% oxygen so that the entire effluent is pure CO2; and there are other possibilities. But all of these require energy to operate, so we will have to use more fossil fuel in order to capture it – also using up fossil fuel reserves more rapidly.

• Once captured, energy will be require to sequester the CO2 somewhere.• One possible reservoir is the deep sea – injection of carbon will eliminate the “lag” of

ocean CO2 uptake, but will leave 20% of emissions in the atmosphere for thousands of years; best thought of as a “peak shaving” device.

• Geological formations are another possible reservoir is; gas and oil have remained trapped in them for 100’s of million years, so they can be quite stable. But the knowledge for reliably pumping CO2 into these reservoirs is not complete, and a lot of geological study and demonstration pilot tests would be necessary before we can rely on containment.

• More speculative solutions: Fe fertilization of Southern Ocean, mirrors in orbit, SO2aerosols in stratosphere.

Vertical chemical profiles in the ocean

Image removed due to copyright restrictions. Citation: Figure 1-1. Broecker, W. S., and T. H. Peng.Tracers in the Sea. Eldigio Press Lamont Doherty Geological Observatory, 1982, 690 pages.

Image removed due to copyright restrictions. Citation: Figure 1-2. Broecker, W. S., and T. H. Peng.Tracers in the Sea. Eldigio Press Lamont Doherty Geological Observatory, 1982, 690 pages.

Documents

12.842 Climate Physics and Chemistry Fall 2006 …dspace.mit.edu/.../0/part2_2pdf.pdfClimate Physics and Chemistry Fall 2006 Natural and Fossil Fuel CO 2 in the Ocean (II) Direct measurements