γλώσσες
Σελίδες
Νομικός
Atmospheric Potential Oxygen (APO) - Distribution, Inversions and
Data NeedsMartin Heimann
with contributions fromChristian Rödenbeck, Corinne LeQuéré, Andrew Manning, Ralph Keeling
Atmospheric Monitoring Network for Biogeochemical Trace Gases (CO2, CH4, CO, N2O, O2/N2, SF6, Isotopes)
CV StationFootprint! !"CV "July
! !q"July
Cape Verde Station2007:
Weekly flask samplinganalysis at MPI-BGC
Late 2007: Addition of continuous measurement system
Max-Planck-Institute for Biogeochemistry
Max-Planck-Society European Union
International Observation Programmes:IGCO, GEOSS, WMO-GAW
CO2 O2 CO2 O2
CO2
O2
Photosynthesis,respiration:
fb ≈ -1.1
Fossil fuel burning:ff ≈ -1.4
Coupling of Carbon and Oxygen Cycles (simplified)
APO = !O2 ! fb!CO2
= !O2 + 1.1 · !CO2
dAPO
dt= !0.3 · QC
foss + QOocean + 1.1 · QC
ocean
Definition of “Atmospheric Potential Oxygen” (APO):
Atmospheric balance equations:
dNC
dt= QC
foss + QCbio + QC
ocean
dNO
dt= ffQC
foss + fbQCbio + QO
ocean
dNO
dt! fb
dNC
dt= (ff ! fb)Q
Cfoss + QO
ocean ! fbQCocean
Properties of the APO Tracer
• Units: ppm (sometimes “permeg”: 1ppm ≈ 4.8 permeg)
• Atmospheric variability: similar to O2 or CO2 variations, i.e. up to a few 10 ppm
• No source/sink over land except small contribution from fossil fuel emissions (-0.3 Qfoss)
• Oceanic sources of APO: contributions from CO2 and from O2 exchanges - time scale dependent
APO = !O2 + 1.1 · !CO2
Measurement Methods for Atmospheric Oxygen
Method Precision O2 vs O2/N2
Interferometry (Keling and Shertz, 1992) ~1 ppm O2/N2
Massspectrometry (Bender, Battle et al.) ~1 ppm O2/N2
Gaschromatography (Toshima et al.) ~3 ppm O2 (?)
Paramagnetic analyser (Manning and Keeling) ~0.2 ppm O2
Vacuum UV absorption (Stephens and Keeling) ~0.2 ppm O2 (?)
Fuel cell (Stephens and Keeling) ~0.5 ppm O2
Main Difficulties:• Small signals (~10-6)• Not absolute standard• Maintenance of accuracy - between sites and with time
APO as an integrator of ocean surface - atmosphere exchange processes
Timescale O2 CO2
Long-term trends (> 5yr) - +++
Longer-term mean spatial patterns
+ +
Interannual variability ++ +
Seasonal cycle +++ +
Synoptic scale +++ -
APO = !O2 + 1.1 · !CO2
Long-term trends: Quantification of the global atmospheric carbon budget
GLOBAL OCEANIC AND LAND BIOTIC CARBON SINKS FROM SCRIPPS NETWORK 103
used the curve fits described in Section 3 to adjust all flask datato the 15th of each month, then monthly means were calculated.For the few months with no flask data, monthly means wereobtained from the curve fits. Twelve consecutive monthly meanswere averaged to compute annual means for APO, with thiscalculation repeated at 6-month time steps centred on 1 Januaryand 1 July of each year.
Table 2 shows annual averages of both O2/N2 ratio data andcalculated APO data for La Jolla, Alert and the average of thesetwo stations. These APO annual averages are shown graphicallyas solid circles in Fig. 3, which also shows calculated annualaverages of CO2 from these two stations. These data points showthe expected trends of decreasing APO over time and increasingCO2 concentrations. Thus the observed atmospheric change atAlert and La Jolla over the 10-yr period from 1990 to 2000 wasan APO decrease of 81.7 ± 8.2 per meg. For !XCO2 (eq. 12), asproposed above, NOAA/CMDL data were used (Conway et al.,1994), thus yielding a land biotic carbon sink that is compatiblewith the more globally representative atmospheric CO2 increaseas determined by the more extensive NOAA/CMDL network.
Oceanuptake
(CO2)global (ppm)354 358 362 366 370 374 378 382 386
OP
Alabolg
)gem rep(
-180
-160
-140
-120
-100
-80
-60
Jan. 1990
July 1998
July 1994
change due to fossil fuelcombustion only
Atmosphericincrease
Land bioticuptake
Jan. 2000
Fig. 3. Vector diagram showing the calculation of the global oceanicand land biotic carbon sinks. Solid circles are annual averages of theobserved APO and CO2 concentrations, calculated by averaging datafrom Alert and La Jolla. Also shown is a fossil fuel combustion line,representing the change in APO and atmospheric CO2 concentrationsthat would have occurred if all CO2 emitted remained in theatmosphere. The slope for the oceanic sink is fixed to an APO:CO2
molar ratio of 1.1 (i.e. "B, see eq. 4), whereas the land biotic sink is ahorizontal line, having no affect on APO, as explained in the text(Section 2). Note that in this figure, CO2 data shown are from our ownmeasurements made at La Jolla and Alert, not from the NOAA/CMDLnetwork used in the IPCC calculations of Section 4.2. This is becausethe NOAA/CMDL data were not in a format that allowed conversion tothis graphical format. However, the purpose of this figure is descriptiveonly. For simplification purposes, this figure does not show the oceanicO2 outgassing term, Zeff.
Table 3. Global fossil fuel combustion data for the 1990s
CO2 O2 APO O2:Cproduceda consumedc consumed molar
Year (Pg C)b (Pg O2) (Pg O2) ratio
1990 6.126 22.648 1.799 1.3881991 6.214 23.091 1.825 1.3951992 6.088 22.605 1.788 1.3941993 6.093 22.657 1.789 1.3961994 6.253 23.214 1.836 1.3931995 6.401 23.715 1.880 1.3911996 6.553 24.330 1.925 1.3941997 6.654 24.662 1.954 1.3911998 6.649 24.747 1.953 1.3971999 6.492 24.217 1.907 1.400Total 63.523 235.886 18.656 1.394d
Total (1014 mol) 52.89 73.71 15.53Total (ppm, per meg) 29.92 198.90 41.91
aData are from Marland et al. (2002), and include CO2 produced fromsolid, liquid and gas fuel, as well as from flared gas and cementmanufacture.b1 Pg is 1015 g, equivalent to 1 Gt.cO2 consumed is calculated assuming full combustion of all fossil fueltypes, and using O2:CO2 molar ratios for each fuel type from Keeling(1988). That is, O2:CO2 is 1.17 for solid fuel; 1.44 for liquid fuel; 1.95for gas fuel; and 1.98 for flared gas. (Cement manufacture does notconsume O2).dAverage O2:CO2 molar ratio.
The global value for !XCO2 was determined to be 15.1 ppmCO2 (equivalent to 32.1 Pg C) using a 2-D atmospheric transportmodel described in Tans et al. (1989).
The amount of CO2 produced from global fossil fuel com-bustion and cement manufacture was calculated from data inMarland et al. (2002). The corresponding amount of O2 (andthus APO) consumed was calculated from a knowledge of therelative fraction of the different fossil fuel types combusted eachyear (Marland et al., 2002) and the average O2:CO2 oxidativeratios for each fuel type given in Keeling (1988a), assumingfull combustion of fossil fuel carbon to CO2. These data areshown in Table 3. For the 10-yr period from January 1990 to Jan-uary 2000 these global fossil fuel emissions resulted in 63.5 ±3.8 Pg C of CO2 being released to the atmosphere (or 52.9 !1014 mol CO2, F in eq. 10; see also bottom of Table 3), and, ifno other processes were involved, would have resulted in a totalAPO decrease of 41.9 ± 2.5 per meg and an atmospheric CO2
increase of 29.9 ± 1.8 ppm (see bottom of Table 3). These hypo-thetical APO and CO2 changes are shown in Fig. 3 as a straightline labelled ‘change due to fossil fuel combustion only’.
The oceanic O2 outgassing term, Z in eqs. 8, 9 and 11, shouldincorporate effects due to both the solubility and biologicalpumps, but in the IPCC calculations repeated here only a sol-ubility correction was applied. A 1 W/m2 warming rate was
Tellus 58B (2006), 2
Manning and Keeling, 2006, Tellus
dNC
dt= QC
foss + QCbio + QC
ocean
dAPO
dt= !0.3 · QC
foss + QOocean + 1.1 · QC
ocean
Small contribution from ocean warming and increased stratification
1993-2003:QCocean = -2.2 ± 0.6 PgCy-1
QCbio = -0.5 ± 0.7 PgCy-1
Long-term mean spatial gradients:
Global ocean carbon model
evaluation
APO
Stephens et al., 1997, GBC
Seasonal cycle:CO2 and O2/N2 Sumburg Head,
Shetland2003-2006
Flask sampling: R. Robertson, Flask measurements: W. Brand, A. Jordan, MPI-BGC
Seasonal cycle relationship:
O2/N2 vs CO2
Sumburgh Head
Slope=-1.7
Seasonal Cycle of APO at Sumburgh Head
Mauna LoaKumukahi
American Samoa
Cape Grim, Tasmania
Palmer Station
South Pole Station
Cold Bay Alaska
Alert Station
Scripps Atmospheric Oxygen Network
Forward and Inverse Modelling
APO source/sink fields
Climate-weather data
Atmospheric transport model
APO concentration
fields
APO concentration observations
Forward Modelling
Comparison
Carbon/oxygen cycle information(GIS, RS, Process
Information)
APO source/sink fields
Climate-weather data
Atmospheric transport model
APO concentration
fields
APO concentration observations
Inverse atmospheric transport modelling
system
Inverse Modelling
Comparison
Carbon/oxygen cycle information(GIS, RS, Process
Information)
• Atmospheric model defines mapping from space of tracer sources to space of tracer mixing ratios χ (concentration C = ρχ):
• Decompose Q(x,t) into series of spatio temporal source-sink patterns:
• Compile a priori sources and uncertainty covariances (qap, Covq)
• Determine αi by minimizing cost function:
!
!t"# = !" · v"# + Q
!mod(x, t) = T · Q(x, t)
Determination of sources and sinks by atmospheric inverse modelling
S2 = (T · q ! !)T· Cov!1
obs · (T · q ! !) + (q ! qap)T· Cov!1
q · (q ! qap)
Q(x, t) =
nq!
i=1
!iqi(x, t)
APO flux model setup
– DRAFT September 15, 2006–
Regions
Example: spatial element
Example: spatial correlation coefficients
Figure 3: Top: Regular grid of regions used in the flux model: Spatially, there is one parameter per suchregion. Colours repeat every 13 regions; in some lines, neighbouring regions happen to have equal colours.Middle: Example of a spatial element g......
Bottom: Spatial correlation coefficients with respect to an example location.
22
Seasonal cycle of APO flues
– DRAFT September 15, 2006–
A
DATA/SOI_MEI_wolters.data
_cg
5 sites (S94A)
Keeling-Garcia & Gruber-Gloor
OCEAN TOTAL
1997 1998 1999 2000 2001 2002 2003-3000
-2000
-1000
0
1000
2000
NH Ocean
-2000
-1000
0
1000
2000
3000
Tropical Ocean
-2000
-1000
0
1000
2000
3000
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1997 1998 1999 2000 2001 2002 2003-3000
-2000
-1000
0
1000
2000
B
DATA/SOI_MEI_wolters.data
_cg
5 sites (S94A)
OPA model
OCEAN TOTAL
1994 1996 1998 2000 2002 2004 2006-400
-300
-200
-100
0
100
NH Ocean
-400
-300
-200
-100
0
100
Tropical Ocean
0
100
200
300
400
500
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1994 1996 1998 2000 2002 2004 2006-500
-400
-300
-200
-100
0
C
DATA/SOI_MEI_wolters.data
_cg
S5CO2 inversion (24 sites)CO2 inversion (40 sites)
CO2 inv. S97ACO2 inv. S94A
OCEAN TOTAL
1994 1996 1998 2000 2002 2004 2006-400
-300
-200
-100
0
100
NH Ocean
-400
-300
-200
-100
0
100
Tropical Ocean
0
100
200
300
400
500
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1994 1996 1998 2000 2002 2004 2006-500
-400
-300
-200
-100
0
Figure 7: Comparison of APO flux estimates to other results. A: Full variability (selected years)compared to flux climatology based on !pO2measurements (Garcia & Keeling, 2001), with long-term values from ocean inversion (Gruber et al., 2001). B: Interannual variations compared to APOfluxes simulated by the OPA ocean process model (?). C: Interannual variations compared to CO2
fluxes estimated from atmospheric data by inversion (Rodenbeck, 2005), multiplied by 1.1.
27
Standard Inversion
Garcia and Keeling, 2001 (O2)+ Gruber et al. (CO2)
20°S-20°N
>20°N
<20°S
Interannual APO fluxes
– DRAFT September 15, 2006–
A
DATA/SOI_MEI_wolters.data
_cg
5 sites (S94A)
Keeling-Garcia & Gruber-Gloor
OCEAN TOTAL
1997 1998 1999 2000 2001 2002 2003-3000
-2000
-1000
0
1000
2000
NH Ocean
-2000
-1000
0
1000
2000
3000
Tropical Ocean
-2000
-1000
0
1000
2000
3000
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1997 1998 1999 2000 2001 2002 2003-3000
-2000
-1000
0
1000
2000
B
DATA/SOI_MEI_wolters.data
_cg
5 sites (S94A)
OPA model
OCEAN TOTAL
1994 1996 1998 2000 2002 2004 2006-400
-300
-200
-100
0
100
NH Ocean
-400
-300
-200
-100
0
100
Tropical Ocean
0
100
200
300
400
500
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1994 1996 1998 2000 2002 2004 2006-500
-400
-300
-200
-100
0
C
DATA/SOI_MEI_wolters.data
_cg
S5CO2 inversion (24 sites)CO2 inversion (40 sites)
CO2 inv. S97ACO2 inv. S94A
OCEAN TOTAL
1994 1996 1998 2000 2002 2004 2006-400
-300
-200
-100
0
100
NH Ocean
-400
-300
-200
-100
0
100
Tropical Ocean
0
100
200
300
400
500
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1994 1996 1998 2000 2002 2004 2006-500
-400
-300
-200
-100
0
Figure 7: Comparison of APO flux estimates to other results. A: Full variability (selected years)compared to flux climatology based on !pO2measurements (Garcia & Keeling, 2001), with long-term values from ocean inversion (Gruber et al., 2001). B: Interannual variations compared to APOfluxes simulated by the OPA ocean process model (?). C: Interannual variations compared to CO2
fluxes estimated from atmospheric data by inversion (Rodenbeck, 2005), multiplied by 1.1.
27
– DRAFT September 15, 2006–
A
DATA/SOI_MEI_wolters.data
_cg
5 sites (S94A)
Keeling-Garcia & Gruber-Gloor
OCEAN TOTAL
1997 1998 1999 2000 2001 2002 2003-3000
-2000
-1000
0
1000
2000
NH Ocean
-2000
-1000
0
1000
2000
3000
Tropical Ocean
-2000
-1000
0
1000
2000
3000A
PO
Flu
x (T
mol
O2/y
ear)
SH Ocean
1997 1998 1999 2000 2001 2002 2003-3000
-2000
-1000
0
1000
2000
B
DATA/SOI_MEI_wolters.data
_cg
5 sites (S94A)
OPA model
OCEAN TOTAL
1994 1996 1998 2000 2002 2004 2006-400
-300
-200
-100
0
100
NH Ocean
-400
-300
-200
-100
0
100
Tropical Ocean
0
100
200
300
400
500
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1994 1996 1998 2000 2002 2004 2006-500
-400
-300
-200
-100
0
C
DATA/SOI_MEI_wolters.data
_cg
S5CO2 inversion (24 sites)CO2 inversion (40 sites)
CO2 inv. S97ACO2 inv. S94A
OCEAN TOTAL
1994 1996 1998 2000 2002 2004 2006-400
-300
-200
-100
0
100
NH Ocean
-400
-300
-200
-100
0
100
Tropical Ocean
0
100
200
300
400
500
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1994 1996 1998 2000 2002 2004 2006-500
-400
-300
-200
-100
0
Figure 7: Comparison of APO flux estimates to other results. A: Full variability (selected years)compared to flux climatology based on !pO2measurements (Garcia & Keeling, 2001), with long-term values from ocean inversion (Gruber et al., 2001). B: Interannual variations compared to APOfluxes simulated by the OPA ocean process model (?). C: Interannual variations compared to CO2
fluxes estimated from atmospheric data by inversion (Rodenbeck, 2005), multiplied by 1.1.
27
– DRAFT September 15, 2006–
A
DATA/SOI_MEI_wolters.data
_cg
5 sites (S94A)
Keeling-Garcia & Gruber-Gloor
OCEAN TOTAL
1997 1998 1999 2000 2001 2002 2003-3000
-2000
-1000
0
1000
2000
NH Ocean
-2000
-1000
0
1000
2000
3000
Tropical Ocean
-2000
-1000
0
1000
2000
3000
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1997 1998 1999 2000 2001 2002 2003-3000
-2000
-1000
0
1000
2000
B
DATA/SOI_MEI_wolters.data
_cg
5 sites (S94A)
OPA model
OCEAN TOTAL
1994 1996 1998 2000 2002 2004 2006-400
-300
-200
-100
0
100
NH Ocean
-400
-300
-200
-100
0
100
Tropical Ocean
0
100
200
300
400
500
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1994 1996 1998 2000 2002 2004 2006-500
-400
-300
-200
-100
0
C
DATA/SOI_MEI_wolters.data
_cg
S5CO2 inversion (24 sites)CO2 inversion (40 sites)
CO2 inv. S97ACO2 inv. S94A
OCEAN TOTAL
1994 1996 1998 2000 2002 2004 2006-400
-300
-200
-100
0
100
NH Ocean
-400
-300
-200
-100
0
100
Tropical Ocean
0
100
200
300
400
500
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1994 1996 1998 2000 2002 2004 2006-500
-400
-300
-200
-100
0
Figure 7: Comparison of APO flux estimates to other results. A: Full variability (selected years)compared to flux climatology based on !pO2measurements (Garcia & Keeling, 2001), with long-term values from ocean inversion (Gruber et al., 2001). B: Interannual variations compared to APOfluxes simulated by the OPA ocean process model (?). C: Interannual variations compared to CO2
fluxes estimated from atmospheric data by inversion (Rodenbeck, 2005), multiplied by 1.1.
27
Standard Inversion
OPA OCCM 20°S-20°N
>20°N
<20°S
Tropical APO fluxes: Links to ENSO?
Standard inversionMEI index (arb. scale)
Corr. coeff ≈ 0.8
– DRAFT September 15, 2006–
A
DATA/SOI_MEI_wolters.data
_cg
S5
S7
S9
OCEAN TOTAL
1994 1996 1998 2000 2002 2004 2006-400
-300
-200
-100
0
100
NH Ocean
-400
-300
-200
-100
0
100
Tropical Ocean
0
100
200
300
400
500
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1994 1996 1998 2000 2002 2004 2006-500
-400
-300
-200
-100
0
Figure 18: As Fig. 6 column A, with ENSO index (MEI) overplotted in red.
38
– DRAFT September 15, 2006–
A
DATA/SOI_MEI_wolters.data
_cg
S5
S7
S9
OCEAN TOTAL
1994 1996 1998 2000 2002 2004 2006-400
-300
-200
-100
0
100
NH Ocean
-400
-300
-200
-100
0
100
Tropical Ocean
0
100
200
300
400
500
AP
O F
lux
(Tm
olO
2/yea
r)
SH Ocean
1994 1996 1998 2000 2002 2004 2006-500
-400
-300
-200
-100
0
Figure 18: As Fig. 6 column A, with ENSO index (MEI) overplotted in red.
38
20°S-20°N
Behrenfeld et al., 2006
approx. 40°S-40°N
Amplitudes:
~ 100 TmolO2 yr-1
~ 200 TmolO2 yr-1!
But NPP is not simply related to APO
O2 undersaturated waters
Equator
Net O2 flux
Foa
Mixed Layer
Latitude
During ENSO event: • reduced upwelling• reduced NPP• reduced O2 uptake• ⇒ anomalous O2 outgassing
But:- C flux changes- Thermal effects
Cape Verde station: Regional signals
Euphotic Zone
ABL
Station
Winddirection
hMBL
Mixing with FT FT
Footprint of Cape Verde station
! !"CV "July
! !q"July
Concluding Remarks
• Atmospheric Potential Oxygen: An integrative tracer of ocean biogeochemistry
• Key drivers:
• NPP, Respiration
• Upwelling
• Temperature
• Global network still very sparse: CV station to provide a key location in the Tropical eastern Atlantic Ocean
• CV Regional Study:
• continuous CO2, O2/N2, and others (N2O,...)
• Flask sampling above ABL (Santo Antao?)
• Flask sampling upstream (Tenerife)
• Surface ocean measurements (IFM-GEOMAR)
• Modelling system (WRF)
Top Related