Southern Ocean Variability

1
Southern Ocean Variability

• Karen J. Heywood
• School of Environmental Sciences
• University of East Anglia
• k.heywood_at_uea.ac.uk
• GEFD
• Summer School
• 2005

2
How do the temperature and salinity of Southern
Ocean water masses vary spatially around
Antarctica? Can we detect temporal changes in
Southern Ocean water masses, such as warming or
freshening? What might cause such changes? How
variable is the transport of the Antarctic
Circumpolar Current? What can we do to improve
our understanding of this variability and its
causes?
3
The Southern Ocean is zonally unbounded. The
Antarctic Circumpolar Current (ACC) transports a
greater volume of water per second than any other
current (about 140 Sverdrups 1 Sv 1 x 106 m3
s-1). Heat, salt, mass, fresh water and
nutrients are exchanged between the Atlantic,
Pacific and Indian Oceans, with consequences for
global climate and ecosystems.
4
Locations of profiles of temperature and salinity
(CTD casts). Small grey dots are all available
historical data. Red dots are those obtained from
the first global snapshot, the World Ocean
Circulation Experiment, 2000-2007.
5
Salinity at 200 m.
Potential temperature at 200 m.
5.5
12.7
0.4
34.12
34.46
35.14
Density of sea water depends on temperature and
salinity, as well as pressure (depth).
6
Neutral density at 200 m depth. By convention,
densities are given with 1000 kg m-3
subtracted. Note the density gradient across the
Antarctic Circumpolar Current. Now imagine that
we follow a particular density surface,
regardless of what depth it is. For example, let
us follow the core density of the Antarctic
Intermediate Water, 27.4.
26.59
27.11
27.95
7
Depth (metres) of the 27.4 isopycnal (neutral
density surface), showing the descent of
Antarctic Intermediate Water. A water mass is
said to form when it leaves the sea surface. It
will carry with it the properties that it had
when it left the surface (for example, the amount
of dissolved gases such as CFCs).
40
600
1050
8
Salinity of the 27.4 isopycnal. This is the
salinity minimum layer denoting Antarctic
Intermediate Water.
33.86
34.22
34.38
9
Potential temperature of the 27.4 isopycnal. As
the Antarctic Intermediate Water penetrates
northward, it mixes with water around it, getting
warmer.
-0.4
3.5
4.5
10
Properties of 28.27 kg m-3 neutral density layer
Antarctic Bottom Water
Potential temperature
Salinity
0.2
34.68
34.69
0.0
0.12
34.71
Antarctic Bottom Water formed in the Ross Sea is
saltier and warmer on density surfaces
(isopycnals) than that formed in the Weddell Sea.
11
Depth (metres) of the 28.27 isopycnal (neutral
density surface). Antarctic Bottom Water
originates on the continental shelves around
Antarctica and descends rapidly to the north.
1800
3500
4500
12

• Temperatures and salinities of Southern Ocean
  water masses vary spatially, according to
• how close to their formation region they are
• how much mixing they have undergone
• which water masses they have mixed with
• how their formation region properties vary.
• We used to think that deep ocean water masses did
  not change temporally. Is this still the case?

13
Five examples of water mass changes detected in
the Southern Ocean in recent decades Warming in
upper 1000 m seen by drifting floats Warming on
density surfaces (at Upper Circumpolar Deep Water
and Antarctic Intermediate Water densities) in
northern Drake Passage Warming on similar
density surfaces in Indian Ocean sector Warming
in Weddell Sea at mid-depths Freshening of
Antarctic Bottom Water off Adelie Land
14
Warming at 900 m depth in the Southern Ocean
Historical data, prior to 1990
Freely drifting floats, after 1990
Floats minus hydrography, showing warming in
Antarctic Circumpolar Current
Gille (2002) Science.
15
Hydrography minus float data for the Southern
Ocean by decade.
Courtesy of Sarah Gille, Scripps
16
Southern ACC
Centre ACC
Northern ACC
Temperature trend (C per decade) across the
Antarctic Circumpolar Current
Courtesy of Sarah Gille, Scripps
17
Elements of the Southern Ocean overturning
circulation
Speer et al. (2000) Journal of Physical
Oceanography
18
Neutral density (kg m-3) across the Southern
Ocean at 30W.
19
Temperature salinity data for the Weddell Sea
and southern Scotia Sea
27.55
Upper Circumpolar Deep Water
28.00
Lower Circumpolar Deep Water
28.26
Weddell Sea Deep Water
28.40
Weddell Sea Bottom Water
Neutral density (?n) contours for the Southern
Ocean
20
Water mass changes in Drake Passage Black Argo
floats 2001-2005 Red 1999 CTD ALBATROSS Red
squares 1990 CTD A21 Green 1980 CTD
Atlantis Yellow, cyan, magenta 1975/6 CTD
ISOS Fronts of the Antarctic Circumpolar
Current SAF Subantarctic Front PF Polar
Front SACCF Southern ACC Front SB Southern
Boundary
21
Deep temperature and salinity differences at
Drake Passage
Antarctic Intermediate Water densities
Upper Circumpolar Deep Water
Upper Circumpolar Deep Water
Lower Circumpolar Deep Water
Courtesy of Dave Stevens, UEA
22
Upper Circumpolar Deep Water
Lower Circumpolar Deep Water
Courtesy of Dave Stevens, UEA
23
Upper Circumpolar Deep Water
Lower Circumpolar Deep Water
Courtesy of Dave Stevens, UEA
24
Upper Circumpolar Deep Water
Lower Circumpolar Deep Water
Courtesy of Dave Stevens, UEA
25
40 years of measurements in Southern Ocean,
Indian sector, by Japanese Antarctic Research
Expedition (JARE)
Syowa
Filled (north) White (south)
Overall warming of 0.005 ?/yr comparable to Gille
(2002) in northern half. Southern half has
larger apparent change maybe because locations of
fronts have shifted. Aoki et al. (2003)
Multi-decadal warming of subsurface temperature
in the Indian sector of the Southern Ocean, J.
Geophys. Res.
Syowa
26
Warming/salinification on 27.9 kg m-3 neutral
density surface in south. Data used JARE
bottle observations with additional Gordon et
al.(1982) data before 1978.
Aoki, Bindoff Church (2005) Interdecadal
watermass changes in the Southern Ocean between
30E and 160E, Geophys. Res. Lett., 32.
27
Schematic diagram of the observed change
Aoki, S., N. L. Bindoff J.A. Church (2005)
Interdecadal watermass changes in the Southern
Ocean between 30E and 160E, GRL, 32.
28
Comparison with the results of HadCM3
anthropogenic run
Cooling
Warming
South pole
North pole
Banks and Bindoff (2003)J.Clim.
29
Long term warming of deep waters of the Weddell
Sea
Maximum potential temperature values of Warm Deep
Water (the temperature maximum layer) in the
Weddell Sea. Robertson et al., 2002
30
Changes detected in Antarctic Bottom Water
WOCE repeated section SR3 and other data
collected at almost the same location and in the
same season. Aoki, S., S.R. Rintoul, S. Ushio,
S. Watanabe N.L. Bindoff, Freshening of the
Adelie Land Bottom Water near 140 E (submitted to
GRL, 2005).
31
Long term freshening of Adelie Land Bottom Water
the variety of Antarctic Bottom Water found in
theAustralian-Antarctic Basin.
Aoki et al., 2005, submitted to GRL.
32

• Likely cause is freshening of source waters
  feeding bottom water, on continental shelf of
  Antarctica.
• Freshening of shelf waters off Adelie Land has
  been observed between 1979 and 2001 (Jacobs,
  2004, Antarctic Science).
• Shelf waters of the Ross Sea also freshening
  (Jacobs et al., 2002, Science).
• Why are the waters on the continental shelf
  around Antarctica getting fresher?
  Possibilities are
• increase in glacial ice melt
• increase in precipitation
• decrease in sea ice production.
• We need to study the processes driving
  variability in the shelf waters proposed for
  the International Polar Year (2007-2009).

33
Increasing evidence that water masses have been
changing during recent decades, especially in
polar regions. Is this a result of
anthropogenic influences ? Or is it natural
climate variability that we have not had the
instrumentation to observe previously? What
might be causing such charges? Need more
climate model studies and more observations! One
possible mechanism is the Southern Annular Mode
(SAM), the main mode of variability in the
atmospheric forcing of the Southern Ocean.
34
See-saw in barometric pressure between Antarctica
and lower-latitudes. Dominant mode of
extra-tropical atmospheric variability in
Southern Hemisphere, on timescales of 10 days
and longer. Thompson Solomon (2002)
Science Marshall et al. (2004) GRL
The Southern Annular Mode

• .
• .

850-hPa height regressed on SAM index
35
Observed sea level pressure trend in ERA40
reanalysis (1955-2005, December-January-February)
Southern Annular Mode index has increased in
recent decades due to increases in both
greenhouse gases and ozone.
Courtesy of Nathan Gillett, UEA
36
Changes in the SAM affect the ocean and
ice. Positive phase of the SAM is predicted to
induce stronger Antarctic Circumpolar Current.
Hall and Visbeck, 2002

37
If the Southern Annular Mode (SAM) has been
increasing, then the forcing driving the
Antarctic Circumpolar Current may have increased.
So has the transport of this current increased in
recent decades? Can we detect any change using
hydrographic sections across the ACC? Using
temperature and salinity sections, we can
calculate the baroclinic geostrophic
current. Using measurements of currents (such as
moored current meters) we can measure the part of
the current that we cannot calculate with
geostrophy (the barotropic part).
38
Transport and variability of the Antarctic
Circumpolar Current in Drake Passage - Repeat
hydrography on WOCE section SR1b
Baroclinic (calculated from temperature and
salinity) transport (Sv) above 3000db
Year mean sd 1975 to 1980 102.7 12.6 1990 to
2000 112.2 5.6 1975 to 2000 107.3 10.4
Courtesy of Stuart Cunningham, NOC
39
Volume transport from the ISOS current meter
moorings during 1979
A moored array of current meters will be deployed
again in Drake Passage in 2006 by Christine
Provost, France.
40
ACC cumulative volume transports relative to the
sea bed south of Australia (SR3) and across Drake
Passage (SR1)
Rintoul Sokolov (2001) Cunningham et al. (2002)
SR3 14710 Sv SR1 137.07.2 Sv
41
Why isnt the volume transport of the ACC the
same at both choke points? Because interbasin
exchanges return the extra water through the
Indonesian archipelago. And there may also be
some transfer to barotropic flow. Courtesy of
Stuart Cunningham, NOC
42
So there is insufficient evidence to determine
whether the transport of the ACC is increasing.
The transport is naturally variable so one
section occupied by a ship once a year is
probably not enough. How else can we measure ACC
variability? Current meter moorings Satellite
altimetry, measuring the slope of the sea
surface across the ACC Tide gauges and pressure
sensors to the south of the ACC?
43
Not correlated
Highly correlated
Correlation of (minus) Drake passage transport
with sea level in a global ocean model suggests
that sea level/bottom pressure close to
Antarctica should be a good indicator of changes
in circumpolar transport on sub-seasonal time
scales.
Hughes, Meredith Heywood, JPO, 1999
44
How can we monitor ocean transport variability?
Tide gauge network on the continental shelf of
Antarctica plus deployments of deep ocean
pressure sensors.
Courtesy of Mike Meredith, BAS
45
OCCAM ¼º model
On sub-seasonal time scales, series are highly
coherent. No discernible lag between forcing
(SAM) and response. No discernible lag around
Antarctica. Strongly related to (model predicted)
Drake Passage transport.

• Hughes et al., GRL, 2003
• Aoki, GRL, 2002

46

• On interannual timescales,
• annual means show significant correlation with
  SAM, and also OCCAM transport, despite baroclinic
  variability being included in the OCCAM value.
• range of transport is quite small (7 Sv).

Meredith et al., GRL, 2004
47
Southern Ocean Variability Water masses vary in
their properties (temperature and salinity)
depending on where they formed and how much they
have mixed. For example, Antarctic Bottom Water
formed in the Ross Sea is saltier than that
formed in the Weddell Sea. There is increasing
evidence that water mass properties in the
Southern ocean are changing. Intermediate depth
waters appear to be warming, consistent with
predictions of global climate models under
anthropogenic forcing. Bottom waters appear to be
freshening, possibly associated with freshening
of shelf waters around Antarctica. The transport
of the Antarctic Circumpolar Current varies by
about 10 and there is no discernible trend.
48
Still Unanswered Questions Why is there warming
of the upper half of the Southern Ocean? Is this
associated with climate change? Are observed
changes due to variations in source water
properties, or shifting of ocean fronts and
current systems? Why are Antarctic shelf waters
freshening? Is this really the cause of bottom
waters freshening, or do we not have enough
observations to eliminate seasonal signals? Is
the ACC transport really stable, or are our
measurements insufficient? Can we use a
combination of sections, moored instruments,
satellite altimetry and models to better
understand and predict ocean variability?