C A S I X

1
C A S I X
Determining Air-Sea fluxes of CO2 from a
synthesis of Earth Observation, coupled models
and in situ measurements Centre for observation
of Air-Sea Interactions and fluXes
Nick Hardman-Mountford, Jim Aiken CASIX Project
Office, Plymouth Marine Laboratory Contributions
from David Woolf (ERI), Peter Challenor, John
Hemmings (NOCS), Rosa Barciela (Met Office),
Helen Kettle (U. Edi.), Paul Monks, Michael
Barkley (U. Leics.), Andy Watson (UEA), Nathalie
Lefevre (LOCEAN) the CASIX Team PML, NOCS/SOES,
POL, UEA, UWB, U.Ply, U.Leics, U.Edi,
U.Read, Met Office
2
Overview

• About CASIX
• Approach
• Work programme
• Examples of key results
• New gas transfer velocity (K) parameterisations
• Satellite data for assimilation/comparison with
  3-D coupled physical-ecosystem models
• In situ observations new instruments and
  interpolation techniques
• Direct CO2 observations from space
• Future

3
CASIX background

• Earth Observation (EO) Centre of Excellence
• Mar 2003- Feb 2008 in last year now
• Focus on ocean-atmosphere CO2 exchange
• Addressing both global ocean and global shelf
  seas
• Primary focus on N Atlantic NW European Shelf
  Seas (validation data available)

4
CASIX purpose exploit EO data to derive air-sea
interactions What are the magnitude, spatial
pattern variability of air-sea CO2 flux?

EO data targets the air-sea interface
X Vertical structure, salinity

• CO2 fluxes driven by physical, chemical
  biological processes, separated in TIME SPACE
• Complexity Models are needed to exploit the
  diverse EO data quantify CO2 fluxes.
• CASIX using 3-D Ocean Shelf circulation models
  with coupled biology (the C-cycle).
• CASIX developing models methods to assimilate
  EO data into 3-D Ocean MODELS.

5
CASIX purpose exploit EO data to derive air-sea
interactions What are the magnitude, spatial
pattern variability of air-sea CO2 flux?

EO data targets the air-sea interface
? Vertical structure, salinity
MODELS 1-D 3-D Ocean and Shelf circulation
coupled biology, C-cycle
6
CASIX science elements and their interaction
1 Physical controls on surface exchange
2 Biogeochemistry and bio-optics
4 Integration (climatology and analysis)
12 Projects in 4 Science Elements 38 Co-Is, RRs,
Collabs 13 Funded (PDRAs) 7 PhDs 3 open now
7
CASIX Major deliverables

• New algorithms for wave breaking and film damping
  from EO data
• Parameterisation of air-sea exchange coefficients
  by EO
• New techniques to estimate primary production
  directly from EO data
• Improved process models of biogeochemical fluxes
  and exchanges
• Tools to assess sensitivity of C flux errors to
  model parameterisations and data assimilation
  procedures
• Algorithms for ocean atmosphere material exchange
  within FOAM
• HadOCC integrated into FOAM ERSEM into POLCOMS
• Operational ocean carbon model, assimilating EO
  ocean colour
• Improved coupled physical-biological shelf seas
  model
• 10 year hind-cast of air sea fluxes for FOAM and
  POLCOMS domains
• 10 year climatologies of air-sea fluxes of CO2
• Analysis of the CO2 climatologies
• Relationships between CO2 fluxes and other
  climate indicators

New EO algorithms
Better understanding of processes
Improved numerical models
CO2 Flux data and climatologies
8
Gas transfer velocity (K)
9
CO2 flux
FCO2 (pCO2(sea) pCO2(atm)) x K x solubility

• Transfer velocity (K)
• K traditionally based on wind speed as proxy for
  sea state
• Sea state available directly from altimeter data
• Wave breaking is a key process

10
Altimeter-based transfer velocity (K)
Wind speed only
Add wave information

• 2 hybrid models dividing K into non-breaking
  (KJ) and breaking wave components
• Empirical uses wind speed only.
• Analytical based on Reynolds number (R) and wave
  height (Hw) - relates to breaking wave properties
  under different fetch regimes
• Woolf (2005), Tellus, 57B, 87-94

11
Dynamic transfer velocity
Spatially seasonally resolved Difference
between Fangohr Woolf (2006) Wanninkhof (1992)
Fangohr Woolf (2006) JMS
12
Application of satellite ocean colour data to
modelled phytoplankton and CO2 fluxes
13
Forecast Ocean Assimilation Model (FOAM)
1º Global
36km (1/3º) North Atlantic and Arctic
12km (1/9º) North Atlantic
6km (1/20º) North East Atlantic
36km (1/3º) Indian Ocean
12km (1/9º) Mediterranean
27km (1/4º) Antarctic

• Met Office operational configurations
• All run daily in the operational suite

12km (1/9º) Arabian Sea
14
Hadley Centre Ocean Carbon Cycle Model (HadOCC)
Aims
- Air-sea fluxes of CO2 using high-res GCM (1º
go, 1/3º 1/9º NA)

• Assimilation of Ocean Colour EO data to
  improve these fluxes

• 10 year hindcast (1997-2006) with/without data
  assimilation

15
Assimilation of derived chlorophyll
Aim Improvement of pCO2 estimation by
assimilating ocean colour

Results from 3-D twin experiments
Phytoplankton background error before the first
analysis.
Phytoplankton analysis error after the first
analysis, with data everywhere.
Phytoplankton errors (mmolN/m3)
16
Daily mean RMS Errors in the North Atlantic
Results from 3-D Twin Experiments

Phytoplankton (mmolN/m3)
Zooplankton (mmolN/m3)
Control - truth
Assimilation - truth
Detritus (mmolN/m3)
Nutrients (mmolN/m3)
Hemmings et al. (in prep)
17
Daily mean RMS Errors in the North Atlantic
Results from 3-D Twin Experiments
Total Dissolved Inorganic Carbon (mmolC/m3)
- Air-sea exchange of CO2 significantly improved
after assimilating ocean colour data
- Joint assimilation of Medspiration SST and
ocean colour is desirable as carbon solubility is
strongly dependent on temperature
- 10 year hindcast will benefit from using a
long-term SST, ocean colour dataset
Hemmings et al. (in prep)
18
Annual cycle assimilating chlorophyll
Free Run
Chlorophyll obs
Barciela et al. (in prep)
19
Annual cycle assimilating chlorophyll
Chlorophyll obs
Assimilation run
Barciela et al. (in prep)
20
Phytoplankton (functional?) types

• Data --- SeaWiFS (Southern summer
• months from 1998 to 2004)
• Method --- absorption from Rrs (IOP approach)
• Blue --- Prokaryotes / Pico eukaryotes
• Yellow--- Flagellates
• Brown --- Diatoms/Dinoflagellates
• Red circle --- South Georgia

Aiken et al. (submitted) Hirata et al. (in prep.)
21

• 12/1998 1/1999
  2/1999 3/1999

12/1999 1/2000
2/2000 3/2000
12/2000 1/2001
2/2001 3/2001
22
12/2001 1/2002
2/2002 3/2002
12/2002 1/2003
2/2003 3/2003
12/2003 1/2004
2/2004 3/2004
23
In situ pCO2 observations and interpolations
24
Comparison of multiple regression and neural
network techniques for mapping in situ pCO2
CAVASOO data Comparison of approaches to
interpolate predict pCO2 based on location,
time and SST. Application to remotely sensed
variables Comparison with modelled fields
Lefèvre et al. (2005), Tellus, 57B, 375-384
25
Systematic errors from SLP averaging

• mean global mass flux (Pg C / yr) computed using
  6 hourly winds and W92 and WM99 for different air
  pressure time averaging periods over 1990-1999
  and for Takahashi et al.'s reference year 1995.
• Net air-sea carbon flux over time.

WM99
W92
Kettle Merchant (2005) ACP
26
pCO2 underway measurement system
Main unit and wet unit in main lab on JCR
See poster by H-M et al.
27
KT supply chain
Remote technical support
Expert advice
NRT processing QC
Daily model validation
Policy-relevant output products
Archiving
Near-autonomousShip pCO2 measurements
See poster by H-M et al.
28
Atmospheric CO2 columns measured from satellite
(SCIAMACHY)
29
(No Transcript)
30
(No Transcript)
31
Barkley et al. (2006) ACP, GRL
32
Summary

• Dynamic K
• sea state variability
• globally seasonally resolved
• pCO2 (sea) and FCO2 fields from models
• constrained by DA of EO parameters (K) and state
  variables (SST, Chl)
• New products (PFTs) to compare with models
• In situ observations essential
• provide validation of models
• new instruments but data still much too sparse
• interpolation techniques provide full fields for
  comparison
• Atmospheric CO2 from satellites
• Best over land

33
Future Work

• Coming year
• New EO and model-based climatologies
• Decadal hindcasts (ocean and shelf)
• Analysis of interannual variability
• Next 5-years
• National Centre for Earth Observation (NCEO)
• C-cycle link ocean, land and atmospheres

34
And thats CASIX (some of, so far).
http//casix.nerc.ac.uk