Mesoscale eddies and deep upwelling in the Southern Ocean Till Kuhlbrodt, Walker Institute, Dept. of Meteorology, NCAS-Climate
Mesoscale eddies in satellite image ½° grid (from Hallberg & Gnanadesikan 2006)
Mesoscale eddies in a 1/6° model Eddy-resolving model (Hallberg & Gnanadesikan 2006)
A simplified zonally averaged section of the Atlantic Drake passage effect Kuhlbrodt et al. (2007), Rev. Geophys. Drake Passage Effect
Antarctic circumpolar current Drake passage effect Antarctic circumpolar current No net zonal pressure gradient no net meridional geostrophic flow Ageostrophic flows: Wind-driven Eddies Drake passage V. Zlotnicki JPL
An outcropping isopycnal layer DOWN W E UP S WIND Eddy activity means downward and upward motion More upwelling than downwelling necessary for a net wind-driven upwelling
The CMIP3 model database
AMOC sensitivity to wind stress Total tx is the integral of the wind stress between 54°S and 64°S (in 1012 N) Numbers from Russell et al. (2006), J. Clim. : observations
Antarctic Circumpolar Current HiGEM CSIRO GISS-AOM GISS-ER GISS-EH Mean ACC strength: (135 ± 84) Sv
AMOC vs. NADW outflow at 30°S Correlation is 0.67 without GISS models “Max. NAOT” is at 50°N
Sensitivity to model resolution Higher resolution better representation of wind-driven upwelling? Higher resolution less numerical diffusion?
Meridional overturning on z-levels Upwelling from depth due to Drake Passage effect (56°S – 62°S) Sinking between 50°S and 56°S Deacon cell (40°S – 60°S) X Kuhlbrodt et al. (2007), Rev. Geophys.
The geometry of the Deacon cell
The invisible Deacon cell The Deacon cell does not appear in a streamfunction plot on potential density levels 1/6° hemispheric model, 20 isopycnal layers (MESO) sinking between 50S and 56S: diapycnal Sigma_2000 necessary to distinguish deep water masses AABW and NADW Red line denotes the zonal maximum time-averaged density of the top 50 m 20 levels, (isopyc), diapycnal diffusivity of 10-5 m2 s-1 Model has strong sponges (restoring) at the equator Hallberg & Gnanadesikan 2006
Mesoscale eddies in a 1/6° model Eddy-resolving model (Hallberg & Gnanadesikan 2006) southward (transient-)eddy-driven flow off the continental shelves importance of three-dimensional picture
Isopycnal streamfunction in OCCAM OCCAM: 1/12°, global model, 66 depth levels (z) Shading: dark – buoyancy gain, light – buoyancy loss Lee et al. 2007 -63
Isopycnal streamfunction in HiGEM Shaffrey et al. 2008, in prep.
Diagnostics Already done: Strength and position of the ACC Meridional streamfunction in isopycnal coordinates Eddy activity Surface wind stress/ Ekman transport Surface buoyancy fluxes What could be looked at: Circulation in individual isopycnal layers Formation of AABW: location, amount Relation to residual mean theory (in zonal average)