A glacial and highresolution Holocene climate record from Law Dome, Antarctica

1
Polynyas, iceshelves,(Antarctic Slope Front,
eddies) and Southern Hemisphere Climate Nathan
Bindoff and others ACE CRC, CSIRO MAR University
of Tasmania, TPAC
2
Polynyas
3
Polynya Processes
4
UP TO 25 FROM AUSTRALIAN ANTARCTIC BASIN Q?
WHERE IS THIS DEEP WATER FORMED
THETA-S VOLUMETRIC CENSUS OF GLOBAL OCEAN BELOW
ZERO CELSIUS Rintoul, Antarctic Res, Ser.,75,
1998
5
CFC-11
6
Mertz Bathymetry
7
Mertz Experiment
8
(No Transcript)
9
Force model with 1990-2000 NCEP/NCAR reanalysis
air temperature, winds, precipitation,
short/longwave radiation to
estimate dense shelf water formation rate, and
its interannual variability
10
Sea Ice Fraction (Aug 1999)
Satellite
Model
Massom et al., Ann. Glaciol., 2001
Marsland et al., JGR, 2004
11
Sea Ice Thickness (Aug 1999)
Satellite
Model
Massom et al., Ann. Glaciol., 2001
Marsland et al., JGR, 2004
12
Circulation below sea-ice at 100m
Ocean circulation driven by buoyancy change
13
Interannual Variability of Dense Shelf Water
Formation

• High Salinity Shelf Water export
• strong interannual variability
• 1991-2000 0.15 Sv
• 1993-1997 0.24 Sv

Marsland et al., JGR, 2004
14
CLIMATE CHANGE
15
SOUTHERN HEMISPHERE SEA ICE VOLUMESENSITIVITY TO
CLIMATE CHANGE
TEMPERATURE
PRECIPITATION

• sea ice decreases with increasing temperature
• sea ice increases with increasing precipitation
• temperature dominates for 2C, 20cm/yr
  freshwater
• 
  Marsland and Wolff, JGR, 2001

16
AIRTEMP -2,-1,1,2 oC PRECIPITATION -20,-10,10,
20 cm/year
MERTZ POLYNYA SENSITIVITY TO CLIMATE CHANGE
SEA ICE EXPORT
-8 1
SEA ICE FRACTION
HEAT FLUX
-8 2
-6 -3
17

What is the Climate Change Response of Dense
Shelf Water Formation?
AIR TEMPERATURE PERTURBATION
Sv
Climate changes in surface forcing lead to small
percentage changes in ice formation, heat flux
etc. But these result in significant cut-off of
dense shelf water production (in short term)!
-25 -39 -82
Sv
PRECIPITATION PERTURBATION
Sv
-19 -29 -61
18
Polynya Conclusions (1)

• Observed outflow is 0.2-0.3 Sv, implying
  0.6-0.9 Sv bottom water formation
• Under climate sensitivity studies
• Approximately 25 reduction of dense water in
  strong polynya years
• Approximately 80 reduction of dense water in
  weak polynya years
• Approximately 40 reduction over all years
• Assuming that the Mertz response is typical of
  other Antarctic coastal polynya regions, then
• expect a slowdown of Southern Hemisphere
  thermohaline circulation
• If future climate tends towards more weak years
  expect a shutdown
• Predict will get the same shutdown of
  thermohaline circulation previously seen in IPCC
  class models, as their horizontal resolution
  becomes sufficient to resolve the coastal polynya
  processes
• Caveat current interannual variability is as
  strong as IPCC forecast changes to climate (need
  for coupled atmosphere?) May get a much larger
  temperature increase!

19
Polynya Conclusions - Modelling (2)

• Pathways from shelf break to abyss
• Poorly resolved and modelled
• Role of canyons and bottom roughness (Jochen
  Kaempf, Flinders University)
• Bottom Boundary Layer schemes (not too diffusive)
• Obtaining bottom layer patterns that look like
  distributions of bottom tracers.
• No explicit ice shelves or Ice Shelf Water
• Important to have realistic coastal geometries
  (eg Mertz Glacier)
• Important to have realistic sea-floor, coastal
  bathymetry important to retaining high salinity
  shelf waters

20
Ice Shelves
21
AMISOR experiment

• Supply of fresh water to oceans
• Icebergs
• melt

22
Amery Ice Shelf
23
AMERY Seasonal Cycle
24
Amery Ice Shelf
25
Amery Ice Shelf
Hunter and Hemer
26
Mertz Glacier Modelling
Guy Williams
27
Ice Shelf Conclusions (1)

• Ice shelves are very sensitive to melt (ocean
  temperatures)
• Amery IS and Mertz Glaciers currently melt 30
  of total ice thats calving.
• Increased ice melt is important component of S.O.
  stratification.
• Much of the melt is coming from depth lt 400
  metres
• Too much emphasis on deep cavities, more on
  shallower depths
• Focus on whole of Antarctica (climate change)

28
Ice Shelf Conclusions (2)

• Whats missing in oceans ?
• Include the contributions of (increased)
  ice-shelf melt to the S.O. freshwater balance.
• Evidence from freshening of bottom waters (Ross
  Sea, Adelie Land)
• Detemine impacts on S.O. stratification and
  overturning circulation.
• Whats missing on ice shelves
• Static thickness and volume
• Active ice shelves, particularly with ice melt
• Time scales for shelves can be short

29
Climate Differences
Steve Griffies
30
Climate Change Detection
31
Climate Change Detection
32
Climate Change Detection
33
Climate Change Detection
34
HadCM3 1990's- 1960's
35
Anthropogenic simulation HadCM3
95
36
More evidence
Aoki et al, 2005
1960s to 1990s 30E to 180
HadCM3 1990's- 1960's
Banks and Bindoff, 2003
37
Warming of the Southern Ocean
OFarrell Budd, in prep
38
Fly In the ointment!
Mixed layer depths and section lines
39
Observations
40
Zonally averaged differences on density surfaces
41
Time series of zonal averages at sigmat26.7
42
Extremes of modelled mixed layer development
43
Sea Level
Sea level change from altimetry 1993-2003
Steric sea level rise from insitu measurements
(and altimetry) 1993-2003
ARGO add dramatically
44
Detection
Parallel Climate Model
Observations
ACC Simulations
Barnett et al. Science 2005
45
Climate Differences (conclusions) (1)

• SAMW seems to be changing throughout S.O.
• AAIW now fresher (and cooler)
• CDW changed unchanged north of SAF
• CDW warmed (saltier) south of SAF
• AABW some evidence now fresher (?)
• Qualitatively same a climate change mode in
  HadCM3 (also CSIRO Mk3)
• Natural variability, aliasing cannot be ignored

46
Climate Differences (Conclusions) (2)

• Rigorous testing of models against current
  climate
• Tuning of parameters
• Very significant improvement in model quality
• Important to save diagnostics (eg mixed layer
  depths, subducted water mass volumes).

47
Summary
48
Movie 1/81/8 model
49
Antarctic Slope Front
50
Eddies

• Mesoscale eddies, motivation and goals for the
  study.
• The model and dataset.
• Results
• General ocean transport, mass, heat and
  freshwater.
• Eddy heat transport.
• Eddy freshwater transport.
• Conclusions.

51
Mesoscale Eddies

• Eddy transport is due to non-cancelling
  variations in heat/salinity and velocity.
• Globally ubiquitous, and may serve as a critical
  property transport mechanism, particularly across
  the ACC. (De Szoeke Levine,1981 Phillips
  Rintoul, 2000).
• Sparse observational studies, modelling only
  recently achieved high resolution. i.e. Semnter
  and Chervin (1992), McCann et al., (1994), Jayne
  and Marotzke (2002).
• Large scale total heat and freshwater transports
  are insensitive to eddy transports in model
  studies, suggesting a compensating mechanism
  (Drijfhout, 1994).

Topex/Poseidon and European Remote Sensing
satellites Rhines, 2001
52
Study Aims

• To examine the mean large scale circulation and
  property transport in the Southern Hemisphere of
  the TPAC OGCM high resolution seasonless model.
• To calculate and examine the corresponding time
  averaged eddy heat and freshwater transports
  identifying the regions and depths where eddy
  transport is significant.
• To analyse the relationship between the mean and
  eddy property transports.

53
Model and Datasets

• TPAC OGCM, based on MOM 3.0.
• 1/8x1/8 resolution, 24 depth levels.
• Fixed wintertime forcing with Gent and McWilliams
  isoneutral mixing.
• High time resolution and statistics collection
  allows the mean and eddy components of the total
  transport to be calculated.

Average over a year
54
1 Sv
10 Sv
0.88 PW
11 Sv
0.19 Sv
9 Sv
10 Sv
0.49 PW
0.35 PW
145 Sv
1.17 PW
144 Sv
155 Sv
2.20 Sv
1.93 PW
1.43 PW
2.28 Sv
2.26 Sv
Mass transport streamfunction (Sv)
55
40 Sv
10 Sv
20 Sv
18 Sv
Meridional Mass transport streamfunction (Sv)
Meridional Overturning Stream Function
56
Depth integrated absolute eddy northward heat
transport log10Wm-1
57

• Mean flow dominant, gyre interiors and the
  Southern Ocean have little eddy transport.
• Cumulative eddy transports are significant over
  broad scales in the tropical Indian and Pacific
  basins.
• Also in the boundary currents and confluences,
  particularly the Agulhas Retroflection.

Atlantic
Indian
Pacific
Atlantic
Indian
Pacific
Atlantic
Indian
Pacific
Atlantic
Indian
Pacific
Cumulative zonal sum of northward heat transport
58

• The total transport is dominated by the mean flow
  at most latitudes.
• Substantial opposing eddy transport in the Indian
  and Pacific tropics.
• Eddies carry between 20-100 (max 0.43 PW) of
  global heat transport across 37-50S

Northward Heat Transport Components (PW)
59

• Depth intensified heat profile results in little
  eddy heat transport below 1000 m.
• Barotropic eddy transports in the tropics and
  south of 40S.
• The subtropical upper layer (0-400 m) eddy
  transport is strongly southward (gt0.2 PW).
• Due to complex vertical structure or bathymetric
  effects?

Cumulative northward eddy heat transport by depth
(PW)
60

• Eddy and mean heat fluxes are strongly
  anti-correlated.
• Mean flux generally dominates, except at Agulhas
  Retroflection.

Total, mean and eddy ocean to atmosphere heat
flux basin integrals (PW)
61
EAC
Agulhas Current
Southern Ocean
South East Pacific
Equatorial Pacific
Equatorial Indian
Pointwise mean and eddy heat flux (ocean heat
convergence) correlation density (normalised)
62
(No Transcript)
63

• Mean flow again dominates the total transport.
  
• Eddies carry up to 60 (max 0.4 Sv) of global
  freshwater transport across 37-45S. Focused on
  the Agulhas Retroflection.

Northward Freshwater Transport Components (PW)
64

• Again there is relatively little eddy transport
  below 1000 m.
• South of 37S eddy transport is largely to the
  north.
• Baroclinic effects dominates to the north, with
  strong surface layer changes, extending deeper
  and further north than in the case of heat.

Cumulative northward eddy freshwater transport by
depth (PW)
65
Agulhas Current
EAC
Southern Ocean
South East Pacific
Equatorial Pacific
Equatorial Indian
Pointwise mean and eddy freshwater flux (ocean
heat convergence) correlation density (normalised)
66
Eddies - Conclusions

• Generally good representation of mass/heat
  transport. High resolution allows for Agulhas
  and Tasman leakages.
• Eddy transport and flux is greatest in the
  tropics, western boundary currents and
  confluences, and along the SAF.
• Eddy transports have a large length scale in the
  tropics, much shorter elsewhere.
• Most eddy activity is restricted to the upper
  1000 m.
• Anti-correlations between mean and eddy
  divergences for both heat and freshwater fluxes.
  
• Possibly due to regions of strong mean flow
  inducing compensating eddy transports via
  increased baroclinic instability.

67
Eddies- conclusions
68
(No Transcript)