Fridtjof Nansen Medal Lecture

1
Fridtjof Nansen Medal Lecture EGU 1st General
Assembly, Nice 2004 Ocean ventilation and
atmospheric CO2 Andrew Watson1 1School of
Environmental Sciences, University of East
Anglia, Norwich, NR4 7TJ, UK
2
Nansen and the Fram
3
Fridtjof Nansen
The Fram
Roald Amundsen
The Maud
Harald Sverdrup
Walter Munk (with Sverdrup)
4
I Ocean mixing and tracer releases
Walter Munk
Wally Broecker
John Shepherd
Jim Lovelock
Jim Ledwell
5
Outline

• Preamble
• Ocean mixing, tracer release, global overturning
  and energy limitation.
• Glacial interglacial atmospheric CO2 change.
• Southern Ocean mixing
• Glacial overturning circulation and how it may
  have been different how this might affect CO2.

6
Tracer releases and deep ocean mixing

• North Atlantic Tracer Release Experiment
• Thermocline mixing is slow, 10-5 m2s-1

7
Tracer releases and deep ocean mixing

• Brazil Basin TRE
• Mixing near topography is very rapid, 10-3
  m2s-1.

Ledwell, J. R., et al., Nature 403 179-182 (2000)
8
Tracer releases and deep ocean mixing

• Greenland Sea TRE
• Nordic seas mid-water and deep mixing is rapid.

9
Where does the ocean mix?

• Observations (microstructure, tracer release and
  ADCP) have shown
• That diapycnal mixing across the main thermocline
  is very slow (10-5 m2s-1)
• That in the interior away from boundaries and
  topography it is similarly low.
• That mixing associated with deep ocean topography
  is typically 100 times faster.

10
Abyssal recipes (1) and (2)

• The global mixing rate required to sustain the
  abyssal circulation is 10-4 m2s-1.
• The power needed in abyssal mixing to sustain
  20Sv bottom water formation is 2 x 1012W
• This is of the same order as the two possible
  sources of energy deep ocean tides and
  wind-induced energy at depth (each 1012W)

Munk W., (1966) Abyssal Recipes. Deep Sea
Research 13, 707-30 Munk W, Wunsch C., (1998 ).
Abyssal recipes II energetics of tidal and wind
mixing Deep-sea Res Pt I 45 1977-2010
11
Is the global overturning circulation limited by
energy available to mix the deep sea?

• .From scale analysis, the power required is
  Qh??, where
• Q is the overturn - rate of formation of bottom
  water
• ?? is density anomaly involved
• h is the scale height of the stratification
• Hence we might expect Q?? to remain approximately
  constant if density contrast increases,
  overturning rate should decrease.
• But from the vertical advection-diffusion
  balance, Q A.?z/h (where ?z is vertical
  diffusivity, A is ocean area).
• implies that, at global or basin scale, ?z should
  also decrease with increasing ??.

See Nilsson Bostrom and Walin, J. Phys.
Oceanogr. 33 2781-2795 (2003)
12
How might the global overturning change with
climate?

• If past climates produced denser deep water, the
  overturning may have been slower.
• Specifically, in glacial time, if most deep water
  was produced by sea-ice/brine rejection in the
  Antarctic, this would change rates of overturning
  and mixing.
• Could be important particularly in helping to
  explain atmospheric CO2 change.

13
II. Glacial-interglacial atmospheric CO2
14
Glacial-interglacial CO2 change

• 80ppm change occurs in 6000 years.
• Change is co-incident with increase in Antarctic
  temperature
• Before Northern Hemisphere ice melt.
• Suggests that Southern Ocean processes are
  largely responsible.

d
a
t
a

-

D
o
m
e

C

D

d
J
o
u
z
e
l

e
t

a
l
.


2
0
0
1

D
o
m
e

C

C
O

2
,
F
l
ü
c
k
i
g
e
r

e
t

a
l
.


2
0
0
2



2
0
0
M
o
n
n
i
n

e
t

a
1

l
.
15
Causes of G-IG atmospheric CO2 change

• After 20 years of research, the cause(s) of
  glacial-interglacial atmospheric CO2 change are
  still unknown.

• Not quite true there is no completely agreed
  explanation, but we do now know a great deal
  about the processes involved.

• The problem is that, while all of the proposed
  mechanisms can be made to work in simple
  conceptual models, no combination has been
  successfully implemented in GCM/climate models.

• One reason is that GCMs are poor at representing
  the overturning circulation (mixing, convection
  and deep water formation are all parameterized,
  not derived from first principles).

16
Causes of Glacial-interglacial CO2 change

• For some processes (for example, temperature
  change, iron fertilization, salinity, growth of
  terrestrial biosphere) we can quantify very
  roughly how much CO2 change they must have
  caused, (and when).

• This leaves a substantial residual still to be
  explained, and only a few candidate mechanisms to
  explain it.
• Candidates are

17
Possible missing mechanisms for CO2 change

• Increased biological productivity of the Southern
  Ocean (eg Sarmiento and Toggweiler, 1984).
• Capping of Polar Southern Ocean by sea ice
  (Stephens and Keeling, 2000).
• Reduced deep ocean ventilation (Toggweiler, 1999)
• Increased stratification of the near-surface
  Southern Ocean.

18
1) Export production at Last Glacial Maximum
Proxies suggest that biological export
productivity in the Polar Southern Ocean was
lower at LGM. Subantarctic export production was
higher. Biological change not the sole answer.
19
2) LGM and present day Southern Ocean ice cover

• LGM winter sea ice

Proxies suggest winter time sea ice was 2x the
area of today. But summer time sea ice was not
very different from today. The sea-ice capping
hypothesis requires gt 90 sea ice coverage all
year.
20
3) And 4) Reduced deep ocean ventilation, greater
stratification?

• Must play a role neither changes in biological
  production nor increased sea ice cover are
  consistent with proxies, unless different
  circulation is also invoked.

• How is Southern component deep water formed and
  ventilated today?.....
• What stratifies the surface today?

21
Schematic of Southern Ocean vertical circulation.

Subantarctic front
Polar front
Low salinity
Eddy flux
Ekman flux
Brine rejection bottom water formation
Brine rejection bottom water formation
AAIW
CDW
NADW
AABW
22
Salinity profiles in the upper 1500m
Near surface polar southern ocean is stratified
by salinity, formed from excess precipitation and
(mostly?) sea ice melting.
23
Deep mixing in the Southern Ocean
Naveiro-Garabato et al., Science 303, 210-213,
2004
24
WOCE A-23 CFCs
Evidence for very rapid mixing in the region of
the Scotia Sea
25
Southern Ocean CFCs overlaid with s0.3
26
Average CFC profiles in the Southern Ocean
CFC 11 (pmol/kg)
Depth (km)

North of 50S
27
Deep mixing in the Southern Ocean

• Order of magnitude of the mixing below 1000 m and
  south of 50S
• CFCs have a time scale in the atmosphere of 20
  years
• CFC profiles have a depth scale gt 1000m
• Kz gt (Zsc)2/2tsc 10-3 m2s-1

(See Haine T. W. N., Watson, A. J., Liddicoat, M.
I., and Dickson, R. R. (1998). The flow of
Antarctic Bottom Water to the Southwest Indian
Ocean estimated using CFCs. J. Geophys. Res. 103
(C12), 27637-27653.)
28
Deep mixing in the Southern Ocean

• AABW is formed largely by brine-rejection
  processes near the Antarctic continent (not open
  ocean convection).
• The Southern Ocean south of 50S is today a zone
  of rapid deep diapycnal mixing.
• Within the Southern Ocean AABW is substantially
  mixed up into the overlying water it leaks
  back towards the surface.
• The AABW found in the rest of the world ocean is
  comparatively much less dense than it would be
  without this intense mixing.

29
Changes in glacial time

• Greater seasonal production of sea ice implies
  more brine rejection, more (or denser) AABW
  formation.
• This also will leave more fresh water at the
  surface greater near-surface stratification?
• The hydrological cycle was weaker
• less intense source of salinity in warm, low
  latitude water
• Less dense water available for deep convection
  following cooling
• AABW becomes the main deep water mass.

30
Schematic of the glacial vertical circulation?.

Subantarctic front
Polar front
Low salinity
Eddy flux
Ekman flux
Brine rejection bottom water formation
Brine rejection bottom water formation
AAIW
AABW
31
Box models and GCMs

• Box type models can reproduce 50-60ppm changes
  due to sea ice, circulation change.
• But GCMs have been unable to do so, so far
• Why? Present day GCMs normally
• Have poor representation of AABW formation
  processes (convection rather than bottom water
  formation)
• Level models tend to be diapycnally leaky
• Isopycnic models have poor representation of
  polar oceans
• Difficult to run for thousands of years as
  necessary in the study of the carbon cycle.

32
Revised circulation box model for atmospheric
CO2 studies i.e after Webb and Suginohara
(2001)
atmosphere
Subantarctic
North Atlantic
Warm surface
AAIW
Polar Southern Ocean
NADW
Deep
Nature 409 37 JAN 4 2001
33
Low-ventilation glacial model (similar to e.g.
Stephens and Keeling (2000), Toggweiler (1999)
atmosphere
Subantarctic
North Atlantic
Warm surface
AAIW
Polar Southern Ocean
Deep
34
Glacial-to-interglacial transition from the model.
Biological production
Atmospheric pCO2 (ppm)
2 x 106
4 x 106
160
180
200
220
240
260
280
Subantarctic
Initial glacial condition
Polar
Increase temperature
Modern reservoir configuration
Modern circulation and productivity
Decrease salinity
Grow terrestrial bosphere
Allow carbonate compensation
35
Conclusions

• Hypothesize that global overturning is
  energy-limited higher density anomaly leads to
  slower overturning.

• The Southern Ocean is today a region of high
  abyssal mixing AABW is extensively modified
  before leaving the Southern Ocean.

• Seasonal sea ice formation helps to stratify the
  polar Southern Ocean, and provides the ultimate
  source of much AABW

• In glacial time, if more seasonal sea ice was
  formed but the hydrological cycle was weaker.
• Rapid brine rejection densifies the deep ocean.
• AABW/AAIW boundary no longer in most intense
  mixing region
• Resulting in lower global overturning rates.
• And an ocean abyssally stratified by salt.

• Box models suggest this could have provided an
  important mechanism to explain the lower
  atmospheric CO2. We await results from GCMs.

36
(No Transcript)