1. Global Ocean Volume Transport Estimation in the 50-year GECCO Assimilation Weiqiang Wang, Armin Köhl, Detlef Stammer Institut für Meerskunde, Zentrum für Meeres- und Kimaforschung, Universität Hamburg Contact: Weiqiang.Wang@zmaw.de SPP – Kolloquium, München, Oct 6-8, 2008

2. Sponsored by Research Grant Application in Priority Programme 1257 “ Mass Transports and Mass Distribution in the Earth System ” (TransOcean) Overall Goal The use of GRACE and GOCE geoid fields will enable the oceanographic community to determine geostrophic barotropic transports and their role in the general circulation. This will enable the community: – Determination of the changing ocean circulation. – Joint evaluation of global sea level, freshwater changes in the ocean, melting of ice. – Determination of steric and eustatic changes in sea level and separation of barotropic and baroclinic currents. – To separate continental hydrography from oceanic mass movement by filtering.

3. WP1 D. Stammer and Armin Köhl (UHH-IFM)  estimates of mass, heat and freshwater transports, their regional and global divergences  estimates of SSH and bottem pressure changes and their relation to mass transport observations.  estimates of a geoid that is consistent with ocean data, ocean dynamics and goid error information. WP2 C. Böning and J. Dengg (IFM-GEOMAR)  compare characteristic patterns in mass distribution associated with different dynamical states of the MOC to those in SSH  examine the resolution dependence of these patterns  early detection of global transport changes in IPCC scenario runs as well as data products from GRACE WP3 M. Visbeck and J. Karstensen (IFM-GEOMAR)  investigation on the characteristics and reasons for transport variability at and between 16°N and 26°N.  analyse SSH from inverted echo sounder data and compare with GRACE geoid based absolute SSH.  analyse SSH from inverted echo sounder data and compare with high and low resolution model. Research Grant Application in Priority Programme 1257 “ Mass Transports and Mass Distribution in the Earth System ” (TransOcean)

4. Model description:  The German partner of the “Estimating the Circulation and Climate of the Ocean” (GECCO) consortium provided a dynamically consistent estimate (1°  1°) of the time-varying ocean circulation over the 50-year period 1952-2001.  combined most of the World Ocean Circulation Experiment (WOCE) observations by using the model’s adjoint to modify the initial temperature and salinity conditions over the full water column and to adjust the time-varying meteorological forcing fields over the full estimation period. Further details on the optimization are provided by Stammer et al. (2004), Köhl et al. (2006, 2007).

5. We are going to show: 1.Sketch of global ocean volume transports 2. Comparison with inverse box model output 3. Decadal variability of global ocean volume transports

6. 1 Sketch of global ocean volume transport For physical consistency in the global ocean, neutral density is used (Jackett and Mcdougall, 1997) and criterions of different layers are as follows(Sun and Rainer, 2001 ) Upper water: (  <27.05) including subtropical Mode water (SMW). Intermediate water: (27.05 <  <27.72) including Antarctic Intermediate Water (AAIW), Subtropical Mode water (SPMW), Subantarctic Mode Water (SAMW) Deep water: (27.72 <  <28.11) including NADW, Circumpolar deep water (CDW) and Labrador Sea water (LSW) Bottom water: (  >28.11) including Lower NADW and Antarctic Bottom Water (AABW). Sketch of global ocean volume transport Depth of 27.05 Depth of 27.72 Depth of 28.11

7. Source from IPCC Ocean conveyor belt Sketch of global ocean volume transport

9. Upper layer red(blue): mass exchange with upper(lower) layer cross(dot): leaving (entering) this layer (  <27.05)

10. Sketch of global ocean volume transport Intermediate layer (27.05 <  <27.72) red(blue): mass exchange with upper(lower) layer cross(dot): leaving (entering) this layer

11. Sketch of global ocean volume transport Deep layer (27.72 <  <28.11) red(blue): mass exchange with upper(lower) layer cross(dot): leaving (entering) this layer

12. Sketch of global ocean volume transport Bottom layer (  <28.11) red(blue): mass exchange with upper(lower) layer cross(dot): leaving (entering) this layer

13. Comparison with inverse box model output 2 Comparison with inverse box model output For the convenience of comparison with Ganachaud and Wunsch (2000, hereafter GW2000), Upper and Intermediate layer (hereafter UI) are analyzed together GW2000 combined WOCE hydrographic data, estimated global circulations and heat flux in an inverse box model. GECCO also combined most of WOCE observations by using the model’s adjoint to modify the initial temperature and salinity conditions over the full water column

14. Comparison with inverse box model output Ganachaud and Wunsch (2000)

15. Deep layer divergence GECCO results: 17Sv of NADW across 32S (Rintoul,1991) 14Sv of deep flow across 24N (Roemmich,1980) Agree very well with observations (134  11 Sv, Cunningham et al. 2003 ). Comparison with inverse box model output The colour of upwelling or downwelling arrows indicates the layer from which the water is coming. ITF transport, consistent with observations (~10 Sv, Gordon et al. 2003 ).

16. GECCO Bottom transport Overturning cell ACC  10 Sv bottom transport convergence (GW2000) in equatorial region of Atlantic, whereas GECCO shows only ~1 Sv. Indian Ocean  GW2000 gives stronger downward transport (21Sv) to bottom layer, which result in much stronger bottom transport in the Atlantic and Indian Ocean than GECCO results. GW2000  GECCOs results are consistent with observations, ~131 Sv in Drake Passage and 141 Sv in section 147E. Whereas GW2000 gives stronger ACC, about 140 Sv in Drake Passage and 157 Sv in section 147E.  GW2000 gives stronger upper/intermediate circulation in India and Pacific Ocean, might associated with upper/intermediate circulation of ACC.  Compared to GECCO result, GW2000 results gives stronger mean overturning circulation not only in Atlantic but also in Indian–Pacific Ocean system.

17. Decadal variability of global ocean volume transport 3. Decadal variability of global ocean volume transport Slowdown of the Atlantic meridional overturning circulation at 25°N (Bryden et al., 2005) from 1957 to 2004 for the layer of 3000-5000m. several decades of measurements are needed before trends can be detected with statistical any significance (Baehr et al., 2007)

18. Decadal variability of global ocean volume transport 43N 25N 30S Atlantic  For the period of 1960-2001, Atlantic MOC is intensified with 1~2 Sv.  The decadal variability of UI and deep layer are in phase for the period 1960-2001.  The bottom transport is minor revision for the whole meridional transport. Numbers marked in figure are the increment during the period

19. Decadal variability of global ocean volume transport Indian & Pacific  One nearly independent overturning cell in Indian and Pacific system which’s combined by ITF.  Unlike in the Atlantic, the transport has no obvious relations between UI, deep and bottom layer.  ITF transport is increasing, which implies the intensified connections between Indian and Pacific Ocean. Indian 34S Pacific 22S ITF Numbers marked in figure are the increment during the period

20. Decadal variability of global ocean volume transport Southern Ocean  ACC is weakening for the period of 1960-2001. The strength of ACC is deceased by 7 Sv in Drake Passage and 20E, around 5 Sv in 147E.  Associated with local decadal variability of deep layer divergences, weakened ACC behaves differently in Drake Passage, section 20E and 147E. UI, deep and bottom transports play different roles in weakened ACC. 20EDrake Passage147E Deep layer divergences in Southern Ocean IO partATL partPAC part

21. Summary:  GECCO assimilation shows that examination of the three-dimensional mass flux field agrees well with a number of global and basin-scale circulation features quantitatively.  Comparison with GW2000, GECCO assimilation shows weaker overturning circulation not only in the Atlantic, but also in Indian-Pacific system produces much less bottom transports because of coarse resolution and ignoring sea-ice. gives consistent ACC transports with observations.  GECCO assimilation provides much time-varying information about global ocean circulation, During the period of 1960-2001, weakened ACC behaves differently in different part of Southern Ocean associated with local deep layer divergences. During the period of 1960-2001, Atlantic MOC is intensified with 1~2 Sv.

22. Research Grant Application in Priority Programme 1257 “ Mass Transports and Mass Distribution in the Earth System ” (TransOcean) WP1 D. Stammer and Armin Köhl (UHH-IFM)  estimates of mass, heat and freshwater transports, their regional and global divergences  estimates of SSH and bottem pressure changes and their relation to mass transport observations.  estimates of a geoid that is consistent with ocean data, ocean dynamics and goid error information. WP2 C. Böning and J. Dengg (IFM-GEOMAR)  compare characteristic patterns in mass distribution associated with different dynamical states of the MOC to those in SSH  examine the resolution dependence of these patterns  early detection of global transport changes in IPCC scenario runs as well as data products from GRACE WP3 M. Visbeck and J. Karstensen (IFM-GEOMAR)  investigation on the characteristics and reasons for transport variability at and between 16°N and 26°N.  analyse SSH from inverted echo sounder data and compare with GRACE geoid based absolute SSH.  analyse SSH from inverted echo sounder data and compare with high and low resolution model.