WW3/OCN Required field exchange for 3D/2D coupling

[pvelissariou1]

Description

Discussion on the required import/export fields for 2D/3D coupling between WW3 and OCN

Suggestion

Please check the suggested requirements and add/modify as needed the following google document: ww3-exports-ocn-3Dwave-terms
NOTE: right-click on the link to open the doc in a new window or tab.

[josephzhang8]

At the request of @uturuncoglu I have added additional outputs from export from WW3 via UFS so he can check them.

These new outputs (SXX, SXY, SYY) are only available via OLDIO (which I understand is what he's using), and users do not need to do anything to get them when using UFS. Look for 'RSXX', 'RSXY', 'RSYY' in the nc outputs.

If this works, I can add other variables for 3D vortex formalism later.

[pvelissariou1]

@gseroka , @saeed-moghimi-noaa Another issue might be with the names of the export/import fields in WW3. We need to make sure that the names are consistent/conformant with UFS field dictionaries as well as with CF guidelines (OPTION 1). @uturuncoglu changed the exported radiation stress field names in WW3 to work in UFS-Coastal. My understanding for changing/modyfing the ufs field dictionary (OPTION 2) is that it will require more testing as it is used in UFS weather model as well. You might want to bring this issue into the discussion tomorrow.

[josephzhang8]

It's easy for me to change the names on SCHISM side.

[pvelissariou1]

@josephzhang8 Thank you, let's see how the WW3 folks will respond. Could you please comment on the export fields from WW3 I listed in the document. I believe the list is complete and it covers all wave induced terms in the OCN side.

[josephzhang8]

@pvelissariou1
RS* are 2D radiation stress from Longuet-Higgins & Stewart formulation.
There will be many more once we change to vortex formulation.

[gseroka]

Thanks @pvelissariou1 and all.

At the request of @uturuncoglu I have added additional outputs from export from WW3 via UFS so he can check them.

@josephzhang8 did you add these to the Google document?

@uturuncoglu changed the exported radiation stress field names in WW3 to work in UFS-Coastal.

@pvelissariou1 RS* are 2D radiation stress from Longuet-Higgins & Stewart formulation. There will be many more once we change to vortex formulation.

@pvelissariou1 are the radiation stresses (RS* or other) included in the Google document?

[josephzhang8]

Thx @gseroka! I did not know that doc but will review/edit.

[pvelissariou1]

@gseroka Greg, the RS* fields are not the fields exported by WW3/NUOPC, they are SCHISM's internal definitions. We pass the fields (described in the doc) via the ESMF/NUOPC api, for example as eastward_wave_radiation_stress for RSXX stress. The variables for the export fields in the doc need to be discussed.

[pvelissariou1]

@gseroka after your meeting today, please update the doc so I can finalize it for our future development tasks in UFS-Coastal. We are trying to document and organize our issues/tasks in UFS-Coastal the best we can, @janahaddad is leading these efforts (thank you so much Jana). Most likely, as we deal with many different components we will have some kind of documentation for all components. Also we need to keep track on the issues as often after model updates things break on the model component side (e.g., SCHISM the last days). @janahaddad, Jana we might start thinking of documenting what are the "stable" commits of each model component to make sure that UFS-Coastal compiles and runs its regression tests properly. Any advice/suggestions on these?

[janahaddad]

Hey @pvelissariou1, yes I've been thinking about this, and @uturuncoglu has a preferred method that's he's been moving forward with: the idea is to keep track of component-level issues in forks for that model component. For example Ufuk recently created this ROMS fork for this purpose, and he's recently been working on WW3-level issues on https://github.com/oceanmodeling/ww3/issues

Re: documenting status -- Ufuk has also recently updated the wiki page to track which configurations are working and passing RT's. Let's chat briefly about this at Monday's tagup!

[uturuncoglu]

@pvelissariou1 @janahaddad Please just keep in your mind that WW3 has two NUOPC/Cap (old one and the NCAR one) and I think if we need to introduce new fields that needs to be done in new cap (also called mesh cap) since almost all the WW3 applications under UFS is using new cap.

[uturuncoglu]

@janahaddad Yes. It is better to open an issue in component side (in out forks). At this point, I test all the configurations (marked as bold) in the wiki page and all are working except ROMS one that I am looking at this point. I also need to update the wiki page and remove issue links from WW3 configurations. I'll do it soon.

[pvelissariou1]

@uturuncoglu We need to keep track on woking/stable commits of the model components so we can freeze the model components to a specific working commit. That is all I am saying. Let's talk about all these next week.

[uturuncoglu]

@pvelissariou1 Yes. That is my aim. Once ROMS issue is solved. I'll freeze the code in both model and component level and put restriction for direct push. Then, all updates and developments needs to pass those tests and create PR. We are very close.

[gseroka]

@gseroka after your meeting today, please update the doc so I can finalize it for our future development tasks in UFS-Coastal. We are trying to document and organize our issues/tasks in UFS-Coastal the best we can, @janahaddad is leading these efforts (thank you so much Jana).

Hi @pvelissariou1, Hendrik took a look at it and didn't have any immediate revisions. He suggested Jessica Meixner to review it so that EMC is informed and also good with it. I will send it over to Jessica now for her review and CC you.

Also, @josephzhang8 may have some edits (he referenced above) on the Google doc which he can do while we await Jessica's review.

[pvelissariou1]

@[email protected], [email protected], @[email protected], @[email protected], @[email protected], @gseroka, @saeed-moghimi-noaa, @[email protected], @[email protected]

Dear all,
Here at CMMB we have compiled a document the describes the requirements to couple WW3 with OCN for both 2D and 3D configurations. The document was shared with the EMC/WW3 developers for an initial evaluation/approval (1st iteration).
At your convinience, could you please review the document and add your suggestions/modifications? After all inputs have been incorporated, I will finalize the document to be discussed again with the WW3 developers for the next steps.
Here is the link to the document: ww3-exports-ocn-3Dwave-terms

Thank you so much for your help on this matter.
Takis

[josephzhang8]

Thank you @aronroland for your willingness to help. Can someone share the WW3 version with Aron used in UFS so he can take a look? Thx.

[saeed-moghimi-noaa]

@yunfangsun Would you please help @aronroland? Thanks

[pvelissariou1]

update on 04/03/2024

• • Need to follow with the various model developers to get their inputs (especially ROMS and FVCOM)
• At this point, I believe the document contains all necessary fields that need to be exported from WW3. I will check again and verify if anything else is missing (e.g. significant wave height, ...).
• @josephzhang8 will update the doc as needed following his incorporation of the vortex force into SCHISM.

[josephzhang8]

@pvelissariou1 : we are still waiting for you to tell us if the latest changes by Ufuk are good with your tests. I'd like to merge them to our master soon. Thx.

[pvelissariou1]

I am creating a few more test cases and testing right now. Overall I saw no issues so far. I will update in the next couple days.

[josephzhang8]

@uturuncoglu @pvelissariou1

I just sent you a request to look into WW3 variables needed by SCHISM for 3D vortex formulation. I'll update g-doc now. Thx.

[josephzhang8]

@uturuncoglu : any updates on 3D vortex coupling? I can test it once the coding is complete. Thx.

[uturuncoglu]

@josephzhang8 Sorry. I could not make any progress on it. I am in a whole day meeting next two days but it would be nice to setup call to discuss about it.

[janahaddad]

@josephzhang8 would you be available in the next week for a 30-min meeting to strategize with us on how to move forward on this?

[josephzhang8]

Sure. We have a meeting 4pm on Monday May 13?

[janahaddad]

Thanks @josephzhang8, Are you available this Wed at 2pm EDT?

[josephzhang8]

I'll be travelling this week @ EPA workshop until Friday.

…

-Joseph Joseph Zhang Office: (804) 684 7466 Web: schism.wiki

________________________________ From: Jana Haddad ***@***.***> Sent: Monday, May 6, 2024 3:06 PM To: oceanmodeling/ufs-coastal ***@***.***> Cc: Y. Joseph Zhang ***@***.***>; Mention ***@***.***> Subject: Re: [oceanmodeling/ufs-coastal] WW3/OCN Required field exchange for 3D/2D coupling (Issue oceanmodeling/schism#9) [EXTERNAL to VIMS received message] Thanks @josephzhang8<https://github.com/josephzhang8>, Are you available this Wed at 2pm EDT? — Reply to this email directly, view it on GitHub<https://github.com/oceanmodeling/ufs-coastal/issues/32#issuecomment-2096715753>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AFBKNZ5DEMIOXLRXIPBCHG3ZA7ICTAVCNFSM6AAAAABC67YYE6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAOJWG4YTKNZVGM>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[janahaddad]

Ok thanks, let's plan to discuss at the existing meeting on Monday May 13!

cc @saeed-moghimi-noaa

[uturuncoglu]

@DeniseWorthen @josephzhang8 This is the one in SCHSIM side to couple with WW3 using vortex formulation (let me correct if I am wrong). The discussion in this issue could have information related to both model components but I think it would be great if we could only add issues related with SCHSIM in here and use oceanmodeling/WW3#11 for WW3 specific development.

[josephzhang8]

@uturuncoglu : you can find the SCHISM 3D setup for Shinnecock test here:

/work2/noaa/nosofs/yjzhang/UFS-project/ufs-coastal-tests/RUN01b

I only included the changed inputs here; other inputs are same as before. You can simply overwrite. manning.gr3 is no needed for 3D. Once the code is ready we can test with this case first. Thx.