34 th International Lige Colloquium Tracer Methods in

34 th International Liège Colloquium Tracer Methods in Geophysical Fluid Dynamics 6 May 2002 Towards Quantitative Evaluation of Ocean Tracer Model Simulations J. C. Orr 1, K. G. Caldeira 2, K. E. Taylor 3 and the OCMIP Group* 1 LSCE/CEA/CNRS and IPSL (France) 2 LLNL 3 PCMDI / LLNL http: //www. ipsl. jussieu. fr/OCMIP

OCMIP-2 Group • • • • AWI (Bremerhaven, Germany): R. Schlitzer, M. -F. Weirig CSIRO (Hobart, Australia): R. Matear IGCR/CCSR (Tokyo, Japan): Y. Yamanaka, A. Ishida IPSL (LSCE, LODy. C, Paris, France): J. Orr, P. Monfray, O. Aumont, J. -Cl. Dutay, P. Brockmann LLNL (Livermore, CA, USA): K. Caldeira, M. Wickett MIT (Boston, USA): M. Follows, J. Marshall MPIM (Max Planck Institut fuer Meteorologie – Hamburg, Germany): E. Maier-Reimer NCAR (Boulder, CO, USA): S. Doney, K. Lindsay, M. Hecht NERSC (Bergen, Norway): H. Drange, Y. Gao PIUB (Bern, Switzerland): F. Joos, K. Plattner PRINCEton (Princeton, USA): J. Sarmiento, A. Gnanadesikan, R. Slater, R. Key SOC (Southampton Oceanography Centre/ Hadley Center, UK): I. Totterdell, A. Yool UL (University of Liege/University Catholique de Louvain, Belgium): A. Mouchet, E. Deleersnijder, J. -M. Campin PMEL/NOAA (Seattle, USA): J. Bullister, C. Sabine PSU (Penn. State, USA): R. Najjar, F. Louanchi UCLA (Los Angeles, USA): N. Gruber, X. Jin

The OCMIP-2 models differ Resolution Seasonality Boundary conditions Sub-grid mixing Mixed Layer Sea-ice Model Offline/Online

How Good is a Model? • Relative to data • Relative to other models • Skill assessment depends on – Our Objectives (e. g. , mean state vs. variability) – Our Vision • Rose colored glasses • Dark Sunglasses • Clear glasses? – Local, Qualitative Analysis – Global, Quantitative Analysis

OCMIP-2 Simulations • Tracers – CFC-11 and CFC-12 – Natural C-14 and Bomb C-14 – He-3 and He-4 • Carbon – Preindustrial: • Abiotic 3 • Common Biogeochemistry (∑CO 2, Alk, PO 4, O 2, DOM) – Preindustrial to Present – Future (two IPCC scenarios: IS 92 a, S 650) – Sequestration (7 sites, 3 depths, 2 scenarios)

Global Ocean Annual Mean Sea-Air CO 2 Flux in 1995 (mol m-2 yr-1)

Some A Summary Useful Diagram* Statistics • Standard deviations – reference – model • Correlation Coefficient R: Key relationship: • Centered Pattern RMS error: Law of Cosines: • Overall Bias: *Taylor, K. E. , J. Geophys. Res. , 106, D 7, 7183 -7192, 2001

Taylor Plot: Sea-Air CO 2 Flux (1995 Annual Mean, Global Map)

Taylor Plot: Sea-Air CO 2 Flux (1995 Full Global Space-Time Distribution)

Basin Zonal Integrals Annual Mean Sea-Air CO 2 Flux in 1995 (Pg C yr-1 deg-1)

Global Seasonal Zonal Integral Sea-Air CO 2 Flux in 1995 (Pg C yr-1 deg-1)

Sea-Air CO 2 Flux: Pacific Ocean North Pacific: 22 o. N-70 o. N Equatorial Pacific: 22 o. S-22 o. N

Conclusions: Total CO 2 validation • Taylor* Diagram: graphical evaluation of 5 global summary statistics (σdata, σmodel, r, R. M. S. , Bias) – Provides quick, global roadmap – Motivates further evaluation – Not mechanistic • Air-sea CO 2 flux – OCMIP-2 models succeed in terms of zonal mean – OCMIP-2 models fail in terms of zonal and seasonal anomalies *Taylor, K. E. , J. Geophys. Res. , 106, D 7, 7183 -7192, 2001

Taylor Plot for Deep 14 C (below 1000 m)

Natural 14 C (West Atlantic, GEOSECS Section) Some models under-predict 14 C Some models over-predict 14 C

Pacific Ocean WOCE P 16 14 C Some models over-predict 14 C Some models under-predict 14 C Data

CFC-11 in the South Atlantic Ajax Data Some models over-predict CFC uptake Dutay et al. , Ocean Modell. , 2001 Some models under-predict CFC uptake

Anthropogenic CO 2 uptake is correlated with CFC-11 uptake and ∆14 C Global CFC-11 Inventory (1989) Mean ∆14 C below 1000 m

Sabine et al. , 2001 Data-based estimate Simulated 1995 anthropogenic CO 2 column inventories Some models take up a lot of CO 2 in the Southern Ocean Some models take up little CO 2 in the Southern Ocean

Global air-sea anthropogenic CO 2 flux (Pg C yr-1)

Summary of OCMIP 2 anthropogenic CO 2 simulations • Some agreement on global historical CO 2 uptake by the ocean – For the 1980’s – For the 1990’s – IS 92 a – adjusted IS 92 a = 2. 0 ± 0. 4 Pg. C yr– 1 = 2. 5 Pg. C yr– 1 = 2. 2 Pg. C yr– 1 • Some disagreement on regional and future fluxes – 1980’s and 1990’s +/-13% (about the mean) – Year 2100 +/-20% (IS 92 a and S 650) – Year 2300 +/-35% (S 650)

Rationale for CO 2 uptake estimate • Observed natural ∆14 C values are within the range of model results. • Modeled CO 2 uptake is correlated with modeled ∆14 C • Observed CFC-11 concentrations are within the range of model results • Modeled CO 2 uptake is correlated with modeled [CFC-11] • Independently estimated anthropogenic CO 2 inventories are within the range of model results • Therefore, anthropogenic CO 2 uptake by the real ocean is probably within the range of model results. • Ocean CO 2 uptake for the 1980’s = 2. 0 ± 0. 4 Pg. C yr– 1 – IS 92 a 1990’s = 2. 5 Pg. C yr– 1 – Adjusted IS 92 a 1990’s = 2. 2 Pg. C yr– 1

Back-up slides

Efficiency of deep CO 2 injection is also correlated with CFC-11 uptake and ∆14 C Global CFC-11 Inventory (1989) Mean ∆14 C below 1000 m

Global Sea-Air CO 2 Flux in 1995 (Zonal Integral, Annual Mean, Pg C yr-1 deg-1)

Pacific Sea-Air CO 2 Flux in 1995 (Zonal Integral, Annual Mean, Pg C yr-1 deg-1)

Anthropogenic CO 2 vs. CFC-11 Gruber et al. , 2001

Large model differences in the Southern Ocean Simulated 1995 cumulative CO 2 fluxes and inventory

US Groups are anticipating funding • Inverse basis function simulations – Gruber et al. • Automated Model Ocean Diagnosis (Auto. MOD), companion to NOCES, focusing on NASA data products – Caldeira et al.

Next steps • Special OCMIP 2 section in Global Biogeochemical Cycles • IGBP open workshop “Global issues in ocean biogeochemistry” – May or June 2002 – GAIM, CLIVAR, JGOFS, GLOBEC, LOICZ, SOLAS • Eur. OCMIP 3 project (NOCES) – Emphasis on interannual to interdecadal variability – A wide range of model types (5 global ocean models, 1 regional ocean model, 1 inverse atmospheric inverse model) • US proposals expecting funding – Inverse basis function simulations (Niki Gruber) – Automated Model Ocean Diagnosis (Ken Caldeira)

Global Biogeochemical Cycles OCMIP 2 Special Section • An overiew and history of OCMIP – J. Orr and N. Gruber • Comparison of model physics relevant to the carbon Cycle in OCMIP 2 – K. Lindsay (NCAR) • OCMIP 2 evaluation of deep-ocean circulation deduced from 3 He and 4 He simulations – J. Dutay (LSCE) • Efficiency of purposeful CO 2 injection in the deep ocean: comparison of the OCMIP 2 models – J. Orr (LSCE) • Air-sea fluxes and north-south ocean transport of CO 2 and O 2: results from OCMIP 2 – P. Monfray (LSCE) • Comparison of new and export production from the OCMIP 2 models – R. Najjar (Penn State) • Simulations of historical and future anthropogenic CO 2 uptake from OCMIP 2 models – J. Orr (LSCE) • Comparisons of model- and data-based estimates of anthropogenic CO 2 in the oceans – C. Sabine (PMEL)

Proposed IGBP open workshop • • TITLE: Global Scale Issues in Ocean Biogeochemistry WHERE: Ispra in Italy, (host N. Hoepffner at JRC) WHEN: May or June 2002 DURATION: 4 days at least (to give time for plenary discussion) SIZE: 60 (? ) with 20 invited CO-ORGANIZERS: R. Schlitzer & P. Monfray ? PURPOSES : – Address big questions in global scale ocean biogeochemistry based on recent data, process modeling, and synthesis – Bridge the gaps between physics, biogeochemistry, and biology – Bridge the gap between people working on observations, process modeling, and global modeling – Stimulate strong concerted action in global scale research TARGET COMMUNITIES: – CLIVAR, JGOFS, GLOBEC, LOICZ and SOLAS