A fully coupled atmosphere–ocean wave modeling system for the  Mediterranean Sea: interactions and sensitivity to the resolved scales and  mechanisms

Katsafados, P.; Papadopoulos, A.; Korres, G.; Varlas, G.

doi:https://doi.org/10.5194/gmd-9-161-2016

Articles | Volume 9, issue 1

https://doi.org/10.5194/gmd-9-161-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/gmd-9-161-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 9, issue 1

Development and technical paper

|

21 Jan 2016

Development and technical paper |

| 21 Jan 2016

A fully coupled atmosphere–ocean wave modeling system for the Mediterranean Sea: interactions and sensitivity to the resolved scales and mechanisms

P. Katsafados, A. Papadopoulos, G. Korres, and G. Varlas

Download

Final revised paper (published on 21 Jan 2016)
Preprint (discussion started on 27 May 2015)
Supplement to the preprint

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC C1179: 'Effect of waves on the atmospheric circulation', Anonymous Referee #1, 18 Jun 2015
- AC C1907: 'Reply to reviewer', Petros Katsafados, 07 Sep 2015
RC C1430: 'Review of A fully coupled Atmosphere-Ocean Wave modeling system for the Mediterranean Sea: interactions and sensitivity to the resolved scales and mechanisms', Anonymous Referee #2, 21 Jul 2015
- AC C1913: 'Reply to reviewer', Petros Katsafados, 07 Sep 2015
EC C1454: 'Referee comments', Andrew Yool, 23 Jul 2015
- AC C1903: 'Reply to the editor', Petros Katsafados, 07 Sep 2015

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Petros Katsafados on behalf of the Authors (07 Sep 2015) Author's response Manuscript

ED: Referee Nomination & Report Request started (25 Sep 2015) by Andrew Yool

RR by Anonymous Referee #1 (11 Oct 2015)

Suggestions for revision or reasons for rejection

Second review of "a fully coupled Atmospheric-Ocean Wave modeling system (WEW) for the Mediterranean Sea: interactions and sensitivity to the resolved scales and mechanisms"

I feel the paper still needs some adjustments to make it acceptable for publication.
As I have stressed in my first review, the concept of coupling an atmospheric model and ocean wave model is not new. Such a system has been operational at ECMWF since 1998 and the method described here is essentially the same, albeit with a different software implementation. So I'm recommending to replace "With this in mind," (line 22) by "Following and adapting concepts already developed and implemented in large scale numerical weather models and in hurricane simulations,"
P3:, line 21: "Charnock drag coefficient" do you mean the drag coefficient obtained when the sea surface aerodynamical roughness is specified using the Charnock relation (Charnock 1955).
P3, line 24: "decrease in sea surface friction arising from the breaking waves" Do you actually mean that? Later at line 27, it is mentioned that "sea surface roughness were much larger". Shouldn't it read instead: " increase in sea surface friction due to strong wave generation" ( momentum is lost by the atmosphere because more waves have to be generated. A large part of that momentum is then lost to the oceans due to wave breaking)
P4: I would move the new line 4 to line 13 (In a recent ... impact) to after line 30.
P4, line 26: I'm not sure we need the exact resolution of the ECMWF system. These will probably be obsolete by the time this paper is published as ECMWF is due to upgrade its resolutions in the spring of 2016 (28km -> 14km ; 16km -> 9 km).
In your review of work, in the My-Wave project, there was also another project that did the same type of work of coupling an atmospheric model with a wave model. It would be good to quote their contribution.
P7, line 27-28: "using OMP directives" add "only" as it is an interesting point to make that the version of WAM you have used did not use MPI directives (you obviously used them for the coupling). As I believe the ECMWF system should serve as a reference, it would be good to point at this obvious difference with the system based on ECWAM. ECMWF makes full use of MPI directives (as well as OMP) and for this reason, it could be fully included in the IFS as a subroutine. Both approaches have merits no doubts. ECMWF system has the advantage that it avoids global field MPI communication among other things.
A note outside the paper review: ECWAM software is available on request to ECMWF.
P7, line 7: Fig. 1 -> Fig. 7 (?)
P8, line 10: the value 0.01 is the value in WAM cycle-4 and in the study presented, you want to show the impact of the coupling without changing the defaults but it would be good to mention in the discussion that the coupled ECWAM is using a lower value of 0.006 (Bidlot 2012 or IFS documentation) because it was found that the mean Charnock value was a bit too high. A sea state dependent Charnock is beneficial but its mean value might need tuning to return to right (unbiased) wind speed in the coupled system.
P10, line 21: we know the advection time step is set by the CFL criteria but could you elaborate a bit more on the choice of the source term integration time step (I know it's probably a compromise between calling WAM every atmospheric time step (too expensive) and a time step that is a multiple of the advection time step but still small enough to capture the high variability of the ETA model. Note that in the ECMWF system, the relatively large atmospheric time step is used instead to determine how often IFS and ECWAM exchange information.
P13, line 9: if I read you correctly, the validation against altimeter wave heights (please state so) was only done for the high-impact selected case (4-11 January 2012), then if it is so, ERS-2 was not longer available.
P13, line 17: bergeron -> Bergeron
P14, line 7-8: I insist, please add Janssen 2004 as all this was already documented in Peter's book.
P15, line 20:-27: your Figure 14 does indicate that for winds above 22 m/s, there is an apparent levelling-off, even a decrease of the Charnock coefficient. If what is presented is the Charnock as modelled by WAM, there is a better explanation for it. It is exactly what you expect from a sea state dependent Charnock. Those winds are for very young sea states and short fetches (i.e. around the fast turning winds). Those young waves are unable to carry the full stress that a slightly more mature sea state could. The process presented in Donelan et al. (2004) of flow separation and continuous wave breaking is not represented in WAM-cycle 4 physics, and so could not be acting. If you were to sample many more storms of different scales, you would find a cloud of points, some still indicating an increase of Charnock with wind speed (see Bidlot 2012). The apparent limitation of the surface drag for high winds is still an ongoing issue as it has become apparent that more than one mechanism is at play. In future, you might contribute to the debate with your WEW system...
Table 1: add the spectral resolution of WAM.
Figures 12,13,14,15: indicate which period/date(s) they cover.
Figure 15: what are the solid lines? Why are the data points stratified?
Finally, referring to Figure 14, I hinted in my previous review that ECMWF has made some changes to the numerical scheme of ECWAM and this can remain outside the paper correction. I did not mean the physics (i.e change around the source terms) but adjustments that made the calculation more accurate. In the case of Figure 14, the apparent increase in Charnock around winds of 6 m/s can be explained by the lack of frequency resolution in the spectrum at high frequency because of the logarithmic frequency spacing and the choice of cut-off frequency. This deficiency affects how the wind input is determined. There is a fix in ECWAM that relies on extending how the parameterised high frequency wave induced stress is determined to slightly lower frequency (Jean Bidlot personal communication). With the frequent exchange of information between the atmosphere and the wave model, it was also found that time integration scheme for the source terms had to be made fully implicit and modified to account for an update on the calculation of the wave induced stress after the update of the spectrum. This change had a beneficial impact, in particular in cases of rapid decreasing wind speeds when otherwise too large Charnock values could be returned to the atmospheric model.

Hide

RR by Anonymous Referee #2 (27 Oct 2015)

ED: Publish subject to minor revisions (Editor review) (10 Nov 2015) by Andrew Yool

AR by Petros Katsafados on behalf of the Authors (20 Nov 2015) Author's response Manuscript

ED: Publish subject to minor revisions (Editor review) (09 Dec 2015) by Andrew Yool

Thank you for your revisions to your manuscript. I believe that the manuscript is now ready for publication.

There are, however, two minor points that I would like to clear up before releasing the manuscript for typesetting.

1. I will contact referee 1 to clarify the point below. In your response you noted that it was not entirely clear what was meant, so I will check that the referee is satisfied. You do not need to do anything at this point.

"P3, line 24: "decrease in sea surface friction arising from the breaking waves" Do you actually mean that? Later at line 27, it is mentioned that "sea surface roughness were much larger". Shouldn't it read instead: " increase in sea surface friction due to strong wave generation" ( momentum is lost by the atmosphere because more waves have to be generated. A large part of that momentum is then lost to the oceans due to wave breaking)"

2. I would be grateful if you could clarify whether your response to the point below involves any changes to the manuscript. It is quite possible that the manuscript already covers the points that you make, but your response is expansive enough that I just want to be clear.

"P10, line 21: we know the advection time step is set by the CFL criteria but could you elaborate a bit more on the choice of the source term integration time step (I know it's probably a compromise between calling WAM every atmospheric time step (too expensive) and a time step that is a multiple of the advection time step but still small enough to capture the high variability of the ETA model. Note that in the ECMWF system, the relatively large atmospheric time step is used instead to determine how often IFS and ECWAM exchange information."

Apologies for this final delay. I will try to push the manuscript through to typesetting as quickly as I can after these points are clarified.

Thank you for your patience.

Andrew Yool.

Hide

AR by Petros Katsafados on behalf of the Authors (14 Dec 2015) Author's response Manuscript

ED: Publish as is (15 Dec 2015) by Andrew Yool

Short summary

This paper includes the entire steps and processes to develop a two-way fully coupled atmosphere-ocean wave model (WEW) aiming a better description and understanding of the exchange processes near the ocean surface. WEW offers a more realistic representation of the extreme weather and sea state events over the ocean bodies and finally leads in an overall improved simulations.