ICOS Science Conference 2020, Session 17

1. Semi-automated near-real time ocean carbon sink
monitoring: its importance and ICOS achievements
Jamie Shutler, Tom Holding, Ian Ashton
University of Exeter
j.d.shutler@exeter.ac.uk

2. Importanceof the ocean

3. Importance of the ocean

4. Importanceof the ocean
Global carbon budgets
The globalland sink of carbon cannot be measured.
Ocean carbon sink provides a powerful constraintfor
identifyingthe landcarbon sink.
Food security and conservation
Monitoringthe ocean sink also identifies regions and
ecosystems at risk.
Accurate estimates of CO2 absorption providea powerful
constrainton carbon budgets and are needed to inform
policies to motivatesocietal shifts towardsreducing carbon
emissions.

5. Exchange across the air-sea interface
Source: Transfer across the air-sea surface, (2013), Springer.

6. Source: Transfer across the air-sea surface, (2013), Springer.
Exchange across the air-sea interface

7. Consolidated methods for temperatureand
salinity handling within gas flux calculations
Woolf,D. K., et al., (2016)On the calculation of air-sea fluxes of CO2 in the presence of temperature and salinity gradients, JGR.
Goddijn-Murphy, L. et al., (2015)The OceanFlux Greenhouse Gases methodologyforderiving a sea surface climatologyof CO2 fugacity in support
of air–sea gas flux studies. Ocean Science.
Source: Woolf et al., (2016)
~100 μm
In situ measurement depth

8. Consolidated methods for temperatureand
salinity handling within gas flux calculations
Quantifiedthe thermal influences on calculationof air-sea gas fluxes:
Clarified a misconception
More accurate calculation
≠
Reduced accuracy
Woolf,D. K., et al., (2016)On the calculation of air-sea fluxes of CO2 in the presence of temperature and salinity gradients, JGR.
Goddijn-Murphy, L. et al., (2015)The OceanFlux Greenhouse Gases methodologyforderiving a sea surface climatologyof CO2 fugacity in support
of air–sea gas flux studies. Ocean Science.

9. Synergy of satellite, in situ and model re-analysis
data is critical
a) example in situ partial pressure; b) spatiallycomplete
atmosphere-ocean CO2 gas fluxes calculatedusing the in situ
data combinedwith Earth observationdata.
a) b)
Satellite observations
Sea state, wind speed, sea skin temperature, salinity
and ice coverage.
Model re-analysisdata
Atmospheric pressure, air temperature,rain
In situ data
CO2 data, water temperature(SOCAT), atmospheric CO2
Interpolation viaSchuster et al. (2013), Biogeosciences.

10. Toolbox developed for community use:
• Open source license and Python library.
• Net flux tool with traceable land/ocean/basin templates.
• User configurable calculation.
• Extensively verified and version controlled.
• Interactivetutorials,used in IOCCP training school.
• So far used within 13 journal papers,
Shutler, J. D. et al., (2016)Flux Engine:A flexible processing system forcalculating air-sea carbon dioxide gas fluxes and climatologies,
JournalofAtmospheric and OceanicTechnology.
Example globalmean daily flux output
Opensource FluxEngine air-seagas fluxtoolbox (Python)
https://github.com/oceanflux-ghg/FluxEngine
Holding, T., et al., (2020), The FluxEngine air-sea gas flux toolbox: simplified interface and extensions for in situ analyses and
multiple poorly soluble gases Ocean Sciences.
in situ satellite reanalysed difference due to reanalysis
interface
centration
2(gCm3
)
sub-skin
concentration
CO2(gCm3
)
SST
(C)
windspeed
(ms1
)
xCO2
(mol/mol)
fCO2
(atm)
10
20
30
0
5
10
394.0
394.5
395.0
300
400
500
0.15
0.20
0.20
30
0
30
0.004
0.000
0.004
0.00
0.01
difference
(gCm3
)
difference
(atm)
fference
gCm3
)
interface
concentration
CO2(gCm3
)
sub-skin
concentration
CO2(gCm3
)
xCO2
(mol/mol)
fCO2
(atm)
CO2flux
(gCm2
day1
)
394.0
394.5
395.0
300
400
500
0.15
0.20
0.15
0.20
0 5 10 15 20
time (days since 16-10-2013)
0.2
0.0
30
0
30
0.004
0.000
0.004
0.01
0.00
0.01
0.05
0.00
0.05
difference
(gCm3
)
difference
(atm)
difference
(gCm3
)
difference
(gCm2
day1
)
interface
concentration
CO2(gCm3
)
sub-skin
concentration
CO2(gCm3
)
SST
(C)
windspeed
(ms1
)
xCO2
(mol/mol)
fCO2
(atm)
O2flux
2
day1
)
10
20
0
5
10
394.0
394.5
395.0
300
400
500
0.15
0.20
0.15
0.20
0.0
30
0
30
0.004
0.000
0.004
0.01
0.00
0.01
0.00
0.05
difference
(gCm3
)
difference
(atm)
difference
(gCm3
)
fference
2
day1
)
Example analysis of in situ cruisedata

11. Routinecarbon sink monitoring is now possible,
but sustained effort is needed
Computing
High precisionand
accuracy satellite data
In situ and model data
eg
Traceable calculation
FluxEngine:verified,
version controlled,open-
source, standard outputs
Requiredcomponents
- Computing facilities.
- Traceablecalculation software;framework for uncertainties.
- Routine satellitedatacollectionand provision.
- Regularatmospheric in situ and model re-analysisdata.
- Regularin water in situ datacollection and collation.
- Experts to oversee the whole process.

12. Example1: Winterweathercontrolsnet influxof
atmosphericCO2 on the North-westEuropean shelf
Vas Kitidis, Jamie Shutler, Ian Ashton, et al (in-review) Winter weather controls net influx of
atmospheric CO2 on the North-west Europeanshelf, Scientific Reports,9(20153),
https://www.nature.com/articles/s41598-019-56363-5
ICOS UK
ICOS France
ICOS Norway
ICOS Germany
EU CARBOCHANGE
EU JERICO-NEXT

13. Example 2: Consolidatedmethods for temperature and
salinity handlingwithingas fluxcalculations
Results
Oceanic sink is at least 0.37 to 0.44 PgC
larger than previously thought (Shutler
et al., 2019).
Globalocean sink valuefor 2010 is:
3 ± 0.6 PgC (Woolf et al., 2019).
Woolf,D. K., et al. (2019) Key uncertainties in recentair-sea CO2 flux, GlobalBiogeochemical cycles, doi: 10.1029/2018GB006041
Shutler, J. D., et al.(2019)Satellites will address critical science priorities forquantifying ocean carbon, Frontiers in Ecologyand Environment.
doi: 10.1002/fee.2129
Data:Holding
et al. (2019),
Pangaea.

14. Watson,A., SchusterU., Shutler, J. D., et al., (2020)Revised estimates of ocean-atmosphere CO2 flux are consistentwith ocean carbon inventory.
Nature Communications, https://doi.org/10.1038/s41467-020-18203-3
Example 3: Revised estimates of ocean-atmosphere CO2 flux are
consistent with ocean carbon inventory
Revised analysis now accounts for thecool,
salty skin of the ocean (blue line)
Revised analysis agrees well with ocean
interior data (thick black and red lines), in
contrastto earlier analyses (dotted lines).
• Ocean sink is much larger than most ocean carbon models suggest.
• This difference in ocean uptake amountsto about 10 per cent of global fossil fuel emissions.
• Result agrees well with an independentsynthesis of interior ocean hydrographydata.
• Analysis supports the revision of landto ocean riverine flux to around 0.5 Pg C yr-1.
• Work suggests that some revision of the global carbon budget is now required.

15. Watson,A., SchusterU., Shutler, J. D., et al., (2020)Revised estimates of ocean-atmosphere CO2 flux are consistentwith ocean carbon inventory.
Nature Communications, https://doi.org/10.1038/s41467-020-18203-3
Revised analysis now accounts for thecool,
salty skin of the ocean (blue line)
Revised analysis agrees well with ocean
interior data (thick black and red lines), in
contrastto earlier analyses (dotted lines).
• Ocean sink is much larger than most ocean carbon models suggest.
• This difference in ocean uptake amountsto about 10 per cent of global fossil fuel emissions.
• Result agrees well with an independentsynthesis of interior ocean hydrographydata.
• Analysis supports the revision of landto ocean riverine flux of 0.5 Pg C yr-1.
• Work suggests that some revision of the global carbon budget is now required.
• Analysis used climate qualitysatelliteEarth observation and in situ data in synergy.
Example 3: Revised estimates of ocean-atmosphere CO2 flux are
consistent with ocean carbon inventory

16. Advancements: Implications for the global
carbon budget of a stronger ocean sink?
Implications of larger ocean sink
Explanationscould include:
1. Emissions are actuallylarger
2. The land sink is smaller.
3. Large variabilityin land to
ocean riverine flux.
Source: CDIAC; NOAA-ESRL;Houghton and Nassikas2017; Hansis et al 2015; Joos et al 2013;
Khatiwalaet al. 2013; DeVries 2014; Friedlingsteinet al 2019; Global Carbon Budget 2019

17. International Charter for Space
and Major Disasters
Commercial satellitedata could be
used to identify episodic changes
But needs updating to acknowledge
“Long-term man-made climate disaster”
Monitoringthe global and regional ocean carbon sinks is needed to
• Balance the carbon budget
• Monitorthe land carbon sink
• Monitormarine ecosystem health
• Advise policy
• Encourage reductionsin emissions
Routine carbon sink monitoringis now possible, but sustainedeffort is needed.
All monitoring componentsexist, but data productionis not guaranteedand requires
sustained funding.
Eg SOCAT is voluntary.The re-analysisof SOCAT data (v1, v2, v4, v6, v2019, v2020) was
funded by ESA, EU ICOS Ringo and EU ICOS Integral.
Industry could providesome help.
Voluntary
In situ and model data
FluxEngine https://github.com/oceanflux-ghg/FluxEngine
Conclusions
Jamie Shutler, j.d.shutler@exeter.ac.uk

19. Toolbox developed for community use:
• Open source license and Python library.
• Net flux tool with traceable land/ocean/basin templates.
• User configurable calculation.
• Extensively verified and version controlled.
• Interactivetutorials,used in IOCCP training school.
• So far used within 13 journal papers,
Shutler, J. D. et al., (2016)Flux Engine:A flexible processing system forcalculating air-sea carbon dioxide gas fluxes and climatologies,
JournalofAtmospheric and OceanicTechnology.
Example globalmean daily flux output
Opensource FluxEngine air-seagas fluxtoolbox (Python)
https://github.com/oceanflux-ghg/FluxEngine
Holding, T., et al., (2020), The FluxEngine air-sea gas flux toolbox: simplified interface and extensions for in situ analyses and
multiple poorly soluble gases Ocean Sciences.
in situ satellite reanalysed difference due to reanalysis
interface
centration
2(gCm3
)
sub-skin
concentration
CO2(gCm3
)
SST
(C)
windspeed
(ms1
)
xCO2
(mol/mol)
fCO2
(atm)
10
20
30
0
5
10
394.0
394.5
395.0
300
400
500
0.15
0.20
0.20
30
0
30
0.004
0.000
0.004
0.00
0.01
difference
(gCm3
)
difference
(atm)
fference
gCm3
)
interface
concentration
CO2(gCm3
)
sub-skin
concentration
CO2(gCm3
)
xCO2
(mol/mol)
fCO2
(atm)
CO2flux
(gCm2
day1
)
394.0
394.5
395.0
300
400
500
0.15
0.20
0.15
0.20
0 5 10 15 20
time (days since 16-10-2013)
0.2
0.0
30
0
30
0.004
0.000
0.004
0.01
0.00
0.01
0.05
0.00
0.05
difference
(gCm3
)
difference
(atm)
difference
(gCm3
)
difference
(gCm2
day1
)
interface
concentration
CO2(gCm3
)
sub-skin
concentration
CO2(gCm3
)
SST
(C)
windspeed
(ms1
)
xCO2
(mol/mol)
fCO2
(atm)
O2flux
2
day1
)
10
20
0
5
10
394.0
394.5
395.0
300
400
500
0.15
0.20
0.15
0.20
0.0
30
0
30
0.004
0.000
0.004
0.01
0.00
0.01
0.00
0.05
difference
(gCm3
)
difference
(atm)
difference
(gCm3
)
fference
2
day1
)
Example analysis of in situ cruisedata

20. But satellite
observations
offer much more!
Simplifiedview of interactions,
exchange and circulationof CO2
within the ocean, identifying
where satelliteEarth observation
is likely to play a leadingrole in
expandingunderstandingand
capability.
Shutler, J. D., Wanninkhof, R., Nightingale, P. D., Woolf,D. K., Bakker, D. C. E., Watson,A., Ashton, I., Holding, T., Chapron, B., Quilfen, Y., Fairall, C.,
Schuster, U., Nakajima, M., Donlon, C. J., (2019)Satellites will address critical science priorities for quantifying ocean carbon, Frontiersin Ecologyand
Environment.doi: 10.1002/fee.2129