Ralph Keeling Scripps Institution of Oceanography

1
Ralph KeelingScripps Institution of
Oceanography
Global oceanic and land carbon sinks from the
Scripps flask sampling networks
2
Atmospheric CO2 budget
?CO2 F O B
3
Atmospheric CO2 and O2 budgets
?CO2 F O B ?O2 -1.4F 1.1B
4
Atmospheric CO2 and O2 budgets
?CO2 F O B ?O2 -1.4F 1.1B Z
Z
5
(No Transcript)
6
Recent O2 based Carbon budgets
Time Outgas Ocean Land
Period Corr.
Sink Sink Manning (2001) 1990-2000
0.10 1.680.5 1.440.7 IPCC(2001) Keeling
1990-2000 0.28 1.86
0.6 1.260.8 Garcia (2002) Manning
1993-2003 0.48 2.24 0.61 0.510.74 Keel
ing (2005, submitted)
Units Pg C yr-1
7
Recent O2 based Carbon budgets
Time Outgas Ocean Land
Period Corr.
Sink Sink Manning (2001) 1990-2000
0.10 1.680.5 1.440.7 IPCC(2001) Keeling
1990-2000 0.28 1.86
0.6 1.260.8 Garcia (2002) Manning
1993-2003 0.48 2.24 0.61 0.510.74 Keel
ing (2005, submitted)
Units Pg C yr-1
Increase in estimated ocean sink results from
(1) Upwards revision of outgassing correction, as
indicated. (2) Observed O2 loss rate higher over
2000-2003 period.
8
Interannual variations in CO2 O2/N2 and 13C/12C
Correlations between CO2, d13C, and O2 imply land
dominance of variability on El Nino time scales
9
Discussion Dominance of land to interannual
variability also supported by atmospheric
inversions. This is now beyond
dispute. Nevertheless, the smaller oceanic
contribution to variability remains poorly
resolved. All available approaches have
problems CO2 Inversions cant distinguish well
between coastal oceans and land fluxes. 13C/12C
complicated by possible variations in isotopic
fractionation factor with land biota
changes. O2 complicated by interannual
variations in air-sea O2 exchange.
10
Discussion, continued Measurements of O2
nevertheless may prove helpful, by providing a
test of ocean models that predict CO2
variability. The test is realizable via the
tracer APO O2 1.1 CO2
Z Air-sea O2 flux
Interannual variability in APO should reflect
interannual variability in the combined air-sea
CO2 and O2 flux, since interannual variability in
fossil-fuel burning (F) is small.
11
Observed versus Modeled variations in APO
Summary of findings Relatively good
model-to-model agreement. Observations show
2x more variability. If models underestimate
APO variability, do they also underestimate CO2
variability? Needs more work to resolve.
12
Acknowledgements
Charles D KeelingAndrew ManningRoberta
HammeBill PaplawskyGalen McKinley Mick Follows
Corinne LeQuere Christian Roedenbeck Laurent
Bopp
13
Ocean biogeochem. Models MPI Jena model Authors
Buitenhuis, LeQuere, Rodgers Physics
OPA-ORCA Bio model Dynamic Green Ocean
type Forcing daily NCEP Resolution 0.5x2
tropics and poles 2x2
sub-tropics Gas exchange Liss and Merlivat
IPSL model Authors Bopp, Rodgers Physics
OPA-ORCA Bio model Dynamic Green Ocean type
Forcing daily NCEP, mixed boundary
conditions Resolution0.5x2 tropics and poles
2x2 sub-tropics Gas
exchange Wanninkhov (1992) MIT model Authors
McKinley, Follows, Marshall Physics
MITgcm-ECCO Biogeo phosphate light based
export Forcing 12 hr NCEP Resolution 1x1
extra-tropics 0.3x1
tropics Gas exchange Wanninkhov (1992)
14
Global APO changes