Discussion on Ventilation Processes in the Thermocline

1
Discussion on Ventilation Processes in the
Thermocline
Martin Visbeck Department of Earth and
Environmental Science,LDEO, Columbia University
2
Questions

• What are key physical ocean processes that must
  be explicitly represented and/or parameterized in
  order to reduce uncertainty in the climate's
  response to anthropogenic forcing?
• How well do the global 1-degree class of models
  do with these processes?
• Will enhanced resolution resolve problems?
• What sorts of resolution are required for the
  different processes?

3
Factors influencingthermocline ventilation

• Surface buoyancy loss
• Ekman convergence
• Large scale circulation
• Mixing (eddies)

4
Mode Water Formation
15
14
13
12
12
Potential Vorticity x 10-12 (ms)-1 on the sq
26.5 surface according to Lozier et al (1995).
Bermuda is located at the mark. Note the
alignment of PV (right) and geostrophic contours
(left).
Montgomery Streamfunction for the sq 26.5
surface relative to a deep reference level. (ci
m2s-2). Adapted from Lozier (personal comm.)
using Hydrobase climatology.
5
Mode Water Formation Processes

• Surface buoyancy loss
• Lateral exchange by eddy mixing
• Convection into a shear flow

6
Questions

• What are key physical ocean processes that must
  be explicitly represented and/or parameterized in
  order to reduce uncertainty in the climate's
  response to anthropogenic forcing?
• How well do the global 1-degree class of models
  do with these processes?
• Will enhanced resolution resolve problems?
• What sorts of resolution are required for the
  different processes?

7
Resolution dependence of mode water formation
(Czeschel, 2001)
V 2.7 106 km3
V 3.6 106 km3
4/3
1/3
V 3.7 106 km3
V 4.3 106 km3
1/12
data
8
Issues

• High resolution seems to improve the
  representation of the mode waters.
• Can one hope that improved parameterizations of
  eddy processes will improve non eddy resolving
  calculations?
• A good example is the Southern Ocean

9
Example
paper submitted A zonally averaged model of the
meridional overturning in the Southern
Ocean Dirk Olbers and Martin Visbeck
10
Estimate K-Eddy
In the adiabatic interior
Ekman cell given from surface wind stress
isopycnal slope from observations
we can obtain K-Eddy as a function of the
residual circulation
11
Observed Wind Stress
NCEP surface wind stress climatology averaged
along mean path of ACC
12
Averaged Density Field
Density (lines) on salinty (contours)
13
Estimate K-Eddy
In the adiabatic interior
the residual circulation is an unknown function
of the density
we construct it to be the sum of two (partial)
Gaussian functions
14
Construct SO-MOC in density space
AAIW
NADW
NADW
AAIW
Combine two Gaussians functions to give NADW
transport of 10 Sv and AAIW of 15 Sv
15
Construct SO-MOC residual circulation
Implied residual circulation
16
Implied k-eddy
Map of k-eddy (m2 s-1) from observed density
and wind field
17
Estimated K-Eddy profile
K-eddy averaged over the ACC region as a function
of depth and varied AAIW transport
18
Do we have to worry about the vertical shape of
k-eddy ?
We find a 3 fold increase of keddy in the upper
500m This is, (I think) is not used in most
parameterization schemes today. However, we are
getting close to the diabatic regime thus there
in a potential important interaction of the
eddies with the mixed layer (CTP)
19
Questions

• What are key physical ocean processes that must
  be explicitly represented and/or parameterized in
  order to reduce uncertainty in the climate's
  response to anthropogenic forcing?
• How well do the global 1-degree class of models
  do with these processes?
• Will enhanced resolution resolve problems?
• What sorts of resolution are required for the
  different processes?

20
Questions

• What are key physical ocean processes that must
  be explicitly represented and/or parameterized in
  order to reduce uncertainty in the climate's
  response to anthropogenic forcing?
• How well do the global 1-degree class of models
  do with these processes?
• Will enhanced resolution resolve problems?
• What sorts of resolution are required for the
  different processes?

21
A zonally averaged model of the meridional
overturning in the Southern Ocean

• Main objective
• develop a theory that includes a parameterisation
  of the interaction between eddies and the mixed
  layer.- goes beyond simple closures by
  employing an intermediate diabatic layer below
  the well mixed surface layer.
• developes largely analytical solutions.

22
Observed Density Field
WOCE Climatology http//www.dkrz.de/u241046/SACs
erver/SACHome.htm Annual average data set covers
the globe with 1x1 resolution and 45 vertical
layers.
23
Averaged Density Field
Average temperature and salinity along estimated
path of the ACC. Use 3ºC at 200m depth as
criterion for ACC position.
24
Averaged Density Field
Density referenced to the surface (black
lines)Slope of isopycnals in colorPercent of
blocked flow in white contours
25
Construct SO-MOC in density space
AAIW
NADW
Combine two tanh functions to give NADW transport
of 10 Sv and AAIW of 15 Sv
26
Estimated K-Eddy profile
K-eddy averaged over the ACC region as a function
of depth and varied NADW transport