Transcript CT4_overview-gehlen.p+

CarboOcean annual meeting 2009
Sølstrand, October 05-09
CT4: Biogeochemical feedbacks on the
oceanic carbon sink
Marion Gehlen
Laboratoire des Sciences du Climat et de l’Environnement
IPSL, CEA-CNRS-UVSQ, Gif-sur-Yvette, France
L.G. Anderson, G.J. Bellerby, J. Bendtsen, L. Bopp, L. Chou, J.P. Gattuso, R.J. Geider,
C. Heinze, T. Johansen, F. Joos, C. Klaas, J. La Roche, S. Martin, C.M. Moore,
U. Riebesell, J. Segschneider, C. Völker, D. Wolf-Gladrow and others
Figures from Friedlingstein et al. (2006)
INPUT
Earth System
Processes:
physical
biogeochemical
OUTPUT
Physical-chemical feedbacks at high latitudes
The Arctic
1. Study the impact of the change in sea ice cover on the CO2 sources and
sinks in the northern North Atlantic and the Arctic Ocean
2. Assessment of the evolution of the CO2 system in the Arctic Mediterranean
and Nordic Seas over the period from 1990-2060 using a coupled
physical-biogeochemical ocean model
Surface ocean pH reduction over this century
will exhibit large regional variability
Reductions in pH over this century will be greatest in the Arctic
Larger
Change in surface
ocean pH in the
21st Century
Large
Bellerby, et al, 2005.
Wintertime climate scenarios for WArag at the
Røst cold water coral system
Model scenario
PIcntrl: If there had been
no anthropogenic carbon
release
SRESA1B
Coral depth range
Model based on proxies for the CO2 system (specific alkalinity and
scenarios of dpCO2) applied to hydrographic output from the Bergen
Climate Model
W
The Southern Ocean
Investigation of air-sea exchange of CO2 under changing CO2, temperature,
wind stress, precipitation, and ice cover by integration and analysis
of results of a fine resolution circulation model
weakening of the Southern Ocean CO2 sink:
0.08 PgC /yr per decade in response to
increased winds south of 45° (positive SAM
phase) attributed to
(1) to an increase in outgassing of natural
carbon which
(2) overcompensates the increase in the
uptake of anthropogenic CO2
mean SO sink 0.1 to 0.6 PgC/yr
Reaction of Southern Ocean nutrients and phytoplankton to
interannual variability in atmospheric forcing
Method: calculate anomalies by susbtracting mean seasonal cycle, regress anomalies onto
normalized Southern Annular Mode (SAM) index
Physics: higher than usual SAM: equatorward Ekman transports, upwelling -> cooling and
slightly deeper mixed layer depth in band around Antarctica
Is there a biological reaction to increased upwelling of nutrients?
Reaction of Southern Ocean nutrients and phytoplankton to
interannual variability in atmospheric forcing
Nutrients: increased upwelling increases silicate and dissolved iron; iron increase is 1020% of background value
Biology: no clear reaction over most of the Southern Ocean, except in subpolar Pacific:
Other factors (cooling, deeper mixed layer) outweigh increase in growth rate through
nutrients
Völker, Hohn, Losch, Wolf-Gladrow, in preparation
Biogeochemical feedbacks
The concept of ocean carbon pumps
 definition: “An ocean carbon pump is defined as a process that depletes
the ocean surface of SCO2 relative to deep-water SCO2”
Volk and Hoffert (1985)
 relevant quantity = export flux below winter mixed layer
 paradigm: pre-industrial C pump was neutral to air-sea gas exchange
 drivers of change:
climate change (T-effect)
atmospheric CO2 (acidification)
boundary fluxes (external input of nutrients)
Climate change (Temperature effect)
Temperature effect on remineralization
talk on Tuesday afternoon by J. Bendsten
« Temperature dependent remineralisation of organic matter and its
influence on the oceanic CO2-uptake »
Changes in input of external nutrients
1. Process understanding:
ballast hypothesis
CO2 – dust – N2 fixation
2. Future projections: changes in dust input and CO2
talk on Tuesday afternoon by J. Segschneider
« Remote input of nutrients in a changing climate »
QUESTION
Why are deep POC fluxes associated to
mineral fluxes?
Do deep-sea fluxes reflect:
Processes occurring at the source?
=> Relation between POC flux and minerals
in the upper layers of the water column
Processes occurring at depth?
=> No relation between POC flux and minerals
in the upper layers of the water column
Christine Klaas & Dieter Wolf Gladrow
POC REMINERALIZATION MODEL APPLIED
TO CARRYING COEFFICIENTS (ƒi)
POC (z) = S Mi ƒi
(z)
∂ƒi - µ ƒ
=
i
∂z
w
µ z
(- w )
ƒi(z) = ƒMi + ƒ0i *
e
Mi: mass flux of mineral i
µ: remineralization length
w: sinking speed
ƒi: mineral-associated POC
ƒMi: mineral-associated POC
„protected, glued or non
degradable“
ƒ0i: mineral-associated
POC at depth z0
Exponential decay model:
POC coefficient (ƒi) vs.
depth
Data-model comparison
POC flux (mg C m-2 d-1)
0
-100
0
-200
-100
-300
-200
-400
-300
-500
Depth (m)
Depth (m)
0
CaCO3
-600
-700
Opal
-800
100
200
300
-400
-500
-600
R2 = 0.69
-700
Data
-900
-800
-1000
0
0.5
1
POC / mineral (w/w)
1.5
-900
-1000
Model
CONCLUSIONS
-POC and mineral fluxes are also tightly correlated in the
upper water column suggesting an influence of minerals on
export and particle properties in and below the mixed
layer
-Results also support the notion of an asymptotic
value at depth (non labile OM or glue effect)
ESTIMATED RAIN RATIO
~ 120 Tmol Bsi y-1
Tréguer et al. [1995]
PIC/POC at 100m:
<=> ~ 7.2 Pg SiO2
=>
~1.1 Pg C y-1 in CaCO<=>
~ 9.2 Pg CaCO3
3
Lee [2001]
0.1
Corg export at 100 m:
10 Pg y-1
200
160
120
80
40
0
400,000
300,000
200,000
100,000
Striking correlation between
tracer of dust input (DAl),
DFe and N2 fixation.
1.2
0.8
0.4
DIN or DIP (nM)
Very low N2 fixation in S.
Atl. where DFe <50pM and
DIP > 200nM
50
40
30
20
10
0
0
DFe (nM)
DAl (nM)
0
N2 fixation (mol m-2 d-1)
for total and <20m fraction
Trichodesmium
(filaments l-1)
<20 m
200
160
120
80
40
0
nifH (copies l-1)
Total and Filamentous
5
4
3
2
1
0
Total
Surface N2 fixation
(nmol l-1 d-1)
AMT 17
1000
100
10
1
-40
-20
0
20
40
Latitude
LaRoche et al.
The interconnection of C/N cycles
Possible future forcings,
expected direction and
likely response of N2
fixation:
Temperature ↑ ↑
Fe ↑
Light
N2
N2
N↓
O2
T, CO2 ↑
Phytoplankton
Diazotrophs
CO2 ↑ ↑?
P
Fe deposition ↑? ↑
N deposition ↑ ↓?
Thermocline
Exported organic matter
N-loss
Stratification (light) ↑ ↑
OMZ
N:P ~ 16
Internal forcings/feedbacks:
N:P<<16
O2 ↓ → Denitrifcation ↑ → N2 fixation ↑
N2 fixation ↑ → Export ↑ → O2 ↓ ….
Changes in external nutrient fluxes
Dust/Fe
Future changes in dust deposition:
Mahowald et al., 2006 predicts a
scale decline in Fe deposition by
2100
Approach
Run from 1860-2100 forced by observed
and projected atmospheric CO2
Linearly interpolate dust fields from 1860,
2000 and 2100
PISCES OBM [Aumont and Bopp, 2006]
=> Examine response of NPP and FCO2
Tagliabue et al. Biogeosciences [2008]
Changes in external nutrient fluxes
Dust/Fe
Large response in areas directly impacted by reduced Fe deposition
But a retroaction in ‘downstream’ regions
Unused nutrients can fuel additional NPP and drive extra CO2 uptake
Changes in external nutrient fluxes
Dust/Fe
Global impacts
•
Fe deposition
60%
•
NPP
3.3%
•
CEX
5%
change in dominant phyto groups
•
FCO2
3.2%
- 22.8 GtC decrease of ocean sink
 small positive feedback
but:  uncertainty with respect to sign of future changes in dust deposition
 impact on marine ecosystems : N-fixation
Why this low sensitivity to Fe input?
Previous models only had an atm source (no seds) and might therefore
have an unrealistically high sensitivity
Sediment Fe included in PISCES and now also hydrothermal Fe input
CO2 effect on seawater chemistry
1. Export stoichiometry
2. Carbonate production and ballast effect
First order quantification of feedback
Changes in ocean chemistry:
Photosynthetic C drawdown
Export stoichiometry
Pelagic Ecosystem CO2 Enrichment Study (PeECE)
(Bergen May 10 – June 12, 2005)
present CO2
2xCO2
3xCO2
C/N=6.7
C/N=7.9
C/N=8.9
“Assuming the observed response can be extrapolated
to new production systems in the ocean, we calculate
an excess CO2 sequestration potential by the biological
carbon pump of 116 Pg C until 2100”
Riebesell et al. (2007)
Changes in ocean chemistry:
C:N = f(pCO2)
Export stoichiometry
projected (SRES A2)
observed
•
Assumption: Mesocosm results can be
extrapolated to global ocean …
•
UVic Earth system model (Schmittner et al., subm.)
1.8 x 3.6 degree resolution, 19 levels
NPZD + diazotrophs ecosystem, (N,P,C,O2)
C:N=const.
C:N=f(pCO2)
•
Simulations from 1765 to 2100
forced by CO2 emissions (historical+SRES A2)
Oschlies et al. (2008)
Changes in ocean chemistry:
Export stoichiometry
Export production
Cumulative signal
C:N=const.
C:N=f(pCO2)
Int(del EP)
Int(del C storage)
Oceanic carbon storage
Increase in export production: 104 GtC
Direct impact on marine C uptake: + 34 GtC
small negative feedback
Oschlies et al. (2008)
D16.17 “Report on results of an arctic mesocosm
experiment focusing on the competition between
phytoplankton and bacteria”
Increasing C:N in a high CO2 world?
But increasing Corg supply to a DOC limited
system resulted in less NCP and less POC as
bacteria outcompeted autotrophs for nutrients
Changes in ocean chemistry:
CaCO3 production
MODEL APPLICATION
 forcing
1 x pCO2 to 4 x pCO2, 1% increase/year
constant circulation
 simulations
CAL01: calcification and dissolution
dependent on Ωc
CAL02: calcification independent,
dissolution dependent of Ωc
CAL03: calcification and dissolution
constant at pre-industrial values
Gehlen et al. (2007)
Changes in ocean chemistry:
CaCO3 production
production - 27 %
CAL01 - 16 %
CAL02 + 19 %
export - 29 %
black:
CAL01
red:
CAL02
blue:
CAL03
dashed:
control
Difference
between cumulative CO2 uptake:
 CaCO3 PRODUCTION and DISSOLUTION
CAL01 - CAL03: + 5.9 GtC
small negative feedback
 CaCO3 DISSOLUTION only CAL01 - CAL03: + 1.2 GtC
Gehlen et al. (2007)
CONCLUSIONS
CarboOcean allowed to:
 assess the contribution of Arctic Seas to the global marine C cycle
 evaluate the Cant inventory of Arctic Seas
 a forecast of changes in the CO2 system of Arctic seas in response to
environmental changes (sea ice cover, river run-off, ocean acidification …)
 assess the variability of the SO CO2 sink in response to climate change, as well as
to forecast impacts on marine productivity
 explore the potential for changes in the biological pump in response to ecosystem
reorganization driven by external input of nutrients, as well as physical and chemical
changes in the ocean: but a start only …
However : Feedbacks still poorly quantified, mostly negative (biogeochemical)
OUTLOOK
 high latitudes identified as regions of rapid environmental change
what about the tropical ocean ?
 continuous effort on process understanding
new targeted experiments ?
synthesis of existing data …
improved synergy between experimentalists & modellers
critical assessment of processes in terms of impact on air-sea fluxes
 model development: interconnexion of biogeochemical cycles: N & C !!
what might be gained from a negative FB in terms of atm CO2, might be
lost in terms of radiative effect of N2O
So far the feedbacks that were quantified are small … existence of thresholds??
The evolution of the marine CO2 system as seen by Clara BARATANGE
early career scientist at LSCE – born in 2007
‘unhappy’ coccolithophore
pHtot
‘happy’ coccolithophore
1886
2100
Time