1. Water mass transformations in the Indonesian Throughflow
Parameterization in an OGCM of the mixing in the ITF : Effect on Water masses
Ariane Koch-Larrouy, Gurvan Madec, Robert Molcard

2. Introduction Only low latitude passage between two oceans
=> key region for circulation and climate
fresh and cool water flow from Pacific Ocean to Indian Ocean
In situ Observations (Gordon 2005)

3. Cruise Only low latitude passage between two oceans
=> key region for circulation and climate
Cruises to understand variability and characteristics of this flow : JADE, ARLINDO, INSTANT
INSTANT ( ), R. Molcard, A. Atmadipoera at the LOCEAN
(+ 300 Indonesian CTD recovered)

4. Water masses transformation
WODB 2001 data
makassar
banda
ceram
halmahera
the flow is going mostly by the western route with 9 Sv in Makassar and then exits in 3 different passages after recirculates in a banda sea. From the south pacific there is a deep flow that concerns 1 Sv this the dashed arrow. The halmahera eddy retroflecs and partly enters the halmahera sea But we don’t know wether or not there is flux at that depth than can enters the indonesian seas as it has been no measurement of it. But when we look at the temperature salinity diagram we can see that we start with a maximum of salinity that is the signature of the SPSW at about 35.5 psu and in Banda sea and just at the exit of ceram sea the salinity profil homogeneous profil of salinity above 20°C.
So the question is to understand why there is a such homogeneous profil of salinity above 20°C.
Some adection diffusion models calculated that a kz of 1 cm2/s is necessary between the entrance to the banda sea. But this is an integrated value that not tells us if the mixing occurs at the sills or in bassin interiors.
Advection diffusion model -> Kz ~ 1-2 cm2/s
Mixing is necessary
but where ?
Ffield & Gordon 92

5. Objectives Tools WHERE ? OPA-NEMO (OGCM) 1/4th degree WHY ?
open boundaries
tke (wind & shear param)
2d internal tides generation model
Results from tidal model
WHERE ?
WHY ?
because of what phenomena
is the ITF transformed
The questions I want to answer are where the mixing occurs and where is the more efficient ?
And because of what
For that I wil use a configuration at 1/4th degree resolution with open boundaries conditions

6. Results .

7. What do we know ? Where ? Why ? Solve the explicit tides ?
2D internal
tides model
Microstructure measurement
Solve the explicit tides ?
Schiller 2004 & Robertson 2006
Show that internal tides could and must be responsible for the mixing
Did not look precisely at the effect on water masses
Hard to link the breaking to the mixing
0.1 cm2/s
60 cm2/s
Hatayama 2004
Alford et al.
Our strategy :
Parameterization of the effects of the tides
Mixing may occur preferentially above rough topography

8. 2 sinks of the tidal energy
Internal tides
2 sinks of the tidal energy
Internal tide generated
Internal wave drag
Energy transfered to barotropic tides to baroclinic tides
bottom friction

9. Internal Tides Internal wave drag from tidal model Lyard & Le Provost
Internal tide generated
Internal wave drag
bottom friction
Le provost & Lyard 2002
This energy transfer is 20 times more concentrated in the ITF than over the global ocean

Lyard & Le Provost

10. Internal Tides 1.1 TW 0.11 TW ITF = unic region in the world
- 20 times more concentrated than for global ocean

- semi enclosed sea => all the energy is avalaible for dissipation
Internal tide generated
Internal wave drag
1.1 TW
bottom friction
0.11 TW

11. parameterization ITF specificities How much is dissipated ? q
St Laurent 2002
How much is dissipated ? q
Where on the horizontal ?
ITF specificities
E(x,y)
Where on the vertical ?
F(z)

12. parameterization All the energy available for mixing q = 1
St Laurent 2002
q tidal dissipation efficiency
Complex topography, series of semi enclosed sea.
Once generated internal tides remain confined
How much is dissipated ? q
Where on the horizontal ?
E(x,y)
Where on the vertical ?
F(z)
All the energy available for mixing q = 1

13. parameterization How much is dissipated ? q = 1
St Laurent 2002
E(x,y) drag coefficient from tidal model
How much is dissipated ?
q = 1
Where on the horizontal ?
Highly heterogeneous
Maximum of energy in Maluku
and Halmahera Seas
E(x,y)
Where on the vertical ?
F(z)
2 tests
- E(x,y) averaged
- E(x,y) apply locally

14. parameterization How much is dissipated ? F(z)~N q = 1 F(z)~N2
St Laurent 2002
F(z) vertical structure of the energy to be dissipated
How much is dissipated ?
F(z)~N
q = 1
F(z)~N2
Where on the horizontal ?
E(x,y)
Where on the vertical ?
F(z)
T. Gerkema &
P. Bouruet Abertot
Internal tidal model 2D
Maximum of energy in the thermocline

15. Results
Max eroded as in the observation
Improved in all the subassin

16. Results
Max eroded as in the observation
Improved in all the subassin

17. Conclusion parameterization
ITF strong internal tides, trapped in the different semi-enclosed seas
build a parameterization of 3d varying kz
average kz = 1.5 cm2/s, independently agrees with the estimates inferred from observations, suggesting that tides are a major phenomenon for the water masses transformation.
the parameterization improves the water masses characteristics in the different Indonesian seas, suggesting that the horizontal and vertical distributions of the mixing are adequately prescribed.
Role of Dewakang sill and Halmahera and Seram seas in mixing
Un peu trop de bla bla

18. Conclusion parameterization
development of a new parameterization taking into account internal tides in an OGCM specific to Indonesian region that reproduce well the water masses and their transformations.
mixing due to internal tides is a major phenomenon explaining the strong transformation of water masses in the ITF
Inter-annual Variability G70 DRAKKAR, comparison with data (INSTANT, …)
Impact of the mixing on a coupled model. Does it modify the atmospheric convection ?

19. Questions ?