NEMO<->wave coupling: modified surface stress with ln_taw and ln_tauoc

Versions v4.2.x
,
1

Hi everyone,
I have a question regarding the coupling of wave stress in NEMO. As I understand there are two options to do so:
(i) component-wise with ln_taw for u and v stress (on u- and v-grids).
(ii) as fraction based on the normalized wave stress with ln_tauoc (on T-grid).

According to the NEMO manual (v4.2.1, Sec. 6.12.7) the surface stress felt by the ocean \tau_{oc} is the atmospheric wind stress \tau_a minus the stress taken up by the waves \tau_w:
\tau_{oc} = \tau_a - \tau_w

However, checking the source code, I see that for option (i), ln_taw=.true., the stress components are updated as (sbcmod.F90, l.460):
utau(:,:) = utau(:,:) - tawx(:,:) + twox(:,:)
vtau(:,:) = vtau(:,:) - tawy(:,:) + twoy(:,:)

While I assume that tawx,y are components of \tau_w (sent to NEMO via OASIS variables O_Tawx, and O_Tawy), what is stress field twox,y (sent to NEMO via OASIS variable O_Twox, and O_Twoy)?

Activating ln_taw, NEMO assumes to receive all the four components tawx, tawy, twox, and twoy, which - in case only tawx,y or twox,y are coupled - results in a unitialized 2nd field leading to Infinity values in utau and vtau and the termination of NEMO.

On the other hand, for option (ii), ln_tauoc=.true., the manual (Sec. 6.12.7) states that
\tau_{oc} = \tau_a \times \tilde{\tau},
with the “normalized wave stress into the ocean” \tilde{\tau}. In the source code (sbcmod.F90 l.447), \tilde{\tau} is tauoc_wave and directly received from OASIS (as O_WStrf):
utau(ji,jj) = utau(ji,jj) * ( tauoc_wave(ji,jj) + tauoc_wave(ji-1,jj) ) * 0.5_wp
vtau(ji,jj) = vtau(ji,jj) * ( tauoc_wave(ji,jj) + tauoc_wave(ji,jj-1) ) * 0.5_wp
Based on the equations above, I would assume that
\tilde{\tau} = \frac{\tau_a - \tau_w}{\tau_a} = 1 - \frac{\tau_w}{\tau_a}.

Therefore, it seems that NEMO is not simply distinguishing normalized/not normalized and T-grid/UV component-wise between the two options ln_taw and ln_tauoc, but substantially expecting different fields. And especially some field that is not really decribed in the manual…

Can someone resolve my confusion here? Which stresses is NEMO exptecting to receive for the two options?

Thanks in advance!
Robert

2

Hi,

For the first option ln_taw=true, indeed it is not fully clear in the documentation.
We will modify the next manual version.

For this option, 4 fields are transferred by the wave model :

  • tawx and tawy which are the net wave supported stress in x and y directions
  • twox and twoy which are the wave to ocean momentum flux in x and y directions

Those fields should be sent to the NEMO u and v grid via the OASIS namcouple file.

Hope this helps!

3

Hi amoulin,
Thanks for your reply! I’ve also found this definition in the sorce code. However, this is what actually confuses me. To my understanding I would only need one of the two fields as \tau_{oc}=\tau_a -\tau_w.
So either I provide the components of net wave supported stress \tau_w (the stress the waves subtract from the total atmospheric stress), which means that:

utau(:,:) = utau(:,:) - tawx(:,:)
vtau(:,:) = vtau(:,:) - tawy(:,:)

Or I directly provide the wave to ocean momentum flux \tau_{oc} (the residual stress remaining from total atmospheric stress), which would mean that:

utau(:,:) = twox(:,:)
vtau(:,:) = twoy(:,:)

However, providing both in the current source code implementation would yield utau=2*twox, wouldn’t it?
Please enlighten me, where my understanding of the terms and of the (supposed) usage of these fields is wrong.

Thanks!
Robert