Peter Wadhams

1
Greenland Sea chimneys and changes in the
thermohaline circulation - the CONVECTION
project
Peter Wadhams
2
Shelf-slope convection
Shelf-slope convection i.e. polynya. The density
of water on the shelf slope increases due to
winter cooling and ice formation, it sinks
through troughs or canyons and mixes with water
off the shelf. If the result is dense enough, it
will sink to the bottom.
3
Open Ocean convection

• Deep open ocean convection only occurs in a
  handful of places
• Not only does deep convection affect the
  properties of the deep waters, but it also
  affects the carbon cycle and oceanic circulation
  through its role in the thermohaline circulation.
  
• Thermohaline circulation helps maintain the flow
  of warm surface waters to high latitudes, and as
  a result northern Europe has an unusually mild
  climate for its latitude.

4
Thermohaline circulation
The large marine conveyor belt with cold deep
flow and warm surface current. (Source
DKRZ/MPI-Hamburg).
Annual mean surface temperature anomalies, from
NCAR data, relative to zonal averages. There is a
5-10C warm anomaly over NW Europe and the Nordic
Seas (Rahmstorf, S. A., Ganapolski, 1999).
5
Northern Hemisphere sites
Deep water production in the northern hemisphere.
Deep Convection occurs in the Labrador Sea,
Greenland Sea and Gulf of Lion. A less important
and sporadic source of Deep water forms in the
Irminger Sea..
(Marshall Schott 1999)
6

• Why these sites?
• Cyclonic circulation the product of which is an
  upward doming of the isopycnals towards the
  centre of the gyre.
• Strong atmospheric forcing Cooling and sea ice
  formation.
• Weakly stratified waters Less water mass
  modification in the central region before it can
  overturn.

7
Convection in the Greenland Sea
Project CONVECTION EU funded, 12 partners, P.
Wadhams co-ordinator Discovered and mapped
long-lived convective chimneys through 3 winters
and 2 summers Developed salt flux model to
predict onset of convection Of central importance
to possible weakening of Atlantic thermohaline
circulation
8
General circulation of the surface and deep
currents within the Nordic Seas
9
E-W potential temperature section at 75N, from
Jan Mayen data (March 2001).
10
What is the signature of convection?
Intermediate convection
Deep convection
11
Convective features
12
What is a chimney?
A rotating column of water of uniform properties
extending from the surface to great depths (2500
m) Found in the centre of the Greenland Sea gyre
in winter Remarkably long-lived (we have followed
one for 3 years) and stable in location (remains
within a few tens of km of 75N 0W) In summer
becomes capped with warm low-salinity water and
appears like a sub-surface eddy, opening back up
to the surface next winter. First chimneys were
found in Mediterranean in 1970, but survive only
a few days. Greenland Sea possesses only
long-lived chimneys, very recently discovered and
not understood. Could they be main vehicle for
deep convection in winter and for maintaining the
Atlantic thermohaline circulation?
13
This is a chimney
Found on March 2001 research cruise of the RV
Jan Mayen 10 km diameter region of uniform
water, 2500 m deep Located at 75N 0E Picture
shows -1.0 C contour penetrating through -0.9C
warm deep layer
14
Evolution
Core temperature of 75/0 chimney, from surveys
done from 2001 onwards. From top to bottom W01,
S01, W02, S02, WS03.
Map of locations of the chimney during surveys of
winter 2001, summer 2001 and winter 2002,
together with contours of convective depths.
15
Polarstern cruise August 2002. (a)
Vessel-mounted ADCP observations of velocity
field at 150-200 m depth horizon. (b)
Cross-section of velocity field of chimney based
on geostrophic calculations corrected by ADCP
observations at 400 m (after Budéus et al.,
2003)
Float trajectories for 5 floats (March to August
1997). From ESOP2.
16
The Odden ice tongue (last seen 1997)
17
Ice area within Odden box 1979 -1999
Did not develop
18
Ice conditions within the Odden
19
Experiments performed within the Odden
Wavebuoy
Pancake drifting buoy
Frazil Sampling
pancake lifting
20
Classical Salinity increases with depth due to
brine drainage.
Inverse Salinity decreases with depth.
Homogeneous Salinity is almost homogeneous
throughout the pancake.
Rafting Pancake rides up on another pancake.
21
Ice dynamics within the Odden
22
Radarsat image from 17 March 2001
East Greenland Current Ice FY MY ice
Odden Ice Pancake and frazil ice
23
(No Transcript)
24
Drift Trajectories
Drift trajectories of the 3 pancake ice buoys.
All buoys were released in the upper portion of
the image. The colour of the trajectories
correspond to the date and the black stars
correspond to the date labels on the colour bar.
25
(No Transcript)
26

• Summing up
• Ice within the central region of the Greenland
  Sea is locally formed frazil and pancake ice
• Pancake ice formed in the central Greenland Sea
  moves with the wind i.e. in free drift.
• Positive salt flux at the formation point and
  negative salt flux where ice melts

27

• What is the role does sea ice play in deep water
  formation?
• A salt flux model incorporates
• ECMWF wind fields,
• passive microwave ice concentration data,
• buoy data and
• in situ field data
• Output
• Ice thickness, salinity and age
• Daily salt flux for the region

28
The Odden and salt flux
29
SSM/I ice concentration
Ice salinity (model)
Ice age (model)
Ice thickness (model)
30
(No Transcript)
31
Conclusions

• Salt refinement is an important factor in
  preparing surface water for convective
  overturning and the deepening of the mixed layer.
• Sea ice does not represent the whole
  manifestation. Other factors contribute to the
  extent of the deepening of the mixed layer and
  hence the depth of convection in any particular
  year. These include
• the amount of freshwater entering the region
  during the previous summer, as well as
  precipitation,
• the strength of the pycnocline and the salinity
  and temperature of the Atlantic layer
• the lateral movement of water masses may also
  play a role in stabilising or destabilising the
  mixed layer.

32
Conclusions

• Very little is known about these features. We
  need to explore
• whether a chimney is a vehicle for active
  ventilation or is an eddy-like feature which may
  have few exchanges with its environment.
• We need to know how such features are spun up and
  how they deteriorate.
• Most importantly, we need to know how long the
  75/0 chimney is going to survive and by what mode
  it disintegrates.

33
Location of chimney with respect to bathymetry.
Contours show depth of convection.
34
The end of the conveyor belt
Convection in the Greenland Sea is the northern
end of the Atlantic thermohaline circulation,
part of the Great Conveyor Belt