UFS-Coastal Meeting Notes

[saeed-moghimi-noaa]

@pvelissariou1 @uturuncoglu @yunfangsun

All,
Please document meeting agenda and notes here by:

• Add an agenda for meetings a couple of days in advance (whoever calling for the meeting)
• What were accomplished by each person
• What action items for each person

Thanks,

[pvelissariou1]

Meeting 05/30/2023: UFS-Coastal presentation(s) and updates
Attendees: Panagiotis Velissariou, Ufuk Turuncoglou, Yunfang Sun

Please modify/add comments as needed

Items discussed:

1. Presentation for UFS-Coastal/CoastalApp in the Storm Surge Seminar Series.
   This was initially set for June 8th but it has been postponed due to the conflict with the ADCIRC User's meeting.

2. Presentation in the Unifying-Innovations 2003

• Takis will start the presentation draft as soon as possible (by mid June)
• Ufuk will contribute on the ufs-coastal part. Also:
  • By that time SCHISM/CMEPS is scheduled to be functional in ufs-coastal
  • By that time ADCIRC/CMEPS is scheduled to be functional in ufs-coastal
• Will show current and development status as well as results from model coupling through CMEPS.

3. Takis will start testing ufs-coastal using SCHISM as the model component. Ufuk will finalize a few outstanding issues in ufs-coastal and will notify us when the application is ready for testing. Testing platforms will be hera and pworks (YunFang).

4. FVCOM needs to be brought to some UFS/NUOPC standards.

• For example, bring the nuopc folder into the FVCOM folder (done)
• Bring the libs/utilities folders outside the src folder
• Update the libraries in the "libs" folder and make sure that the updated libraries compile and work with FVCOM (a few of these libraries won't be supported by UFS so, they need to be compiled on the fly). Takis pointed out that CoastalApp has this capability built-in now.
• Ufuk will check FVCOM (feature/FVCOM in CoastalApp) to see what else needs to be done to bring FVCOM close to some compatibility with UFS
• While CMake compilation of FVCOM is not absolutely necessary at this stage, it is recommended that FVCOM to have this capability implemented sooner than later.

5. FVCOM developers requested to communicate issues/todo things via email rather than Github issues. Takis will prepare a list of modifications/updates to FVCOM so the developers can see how to start working on those to update FVCOM (some of the suggested changes have already been implemented).

[saeed-moghimi-noaa]

@pvelissariou1

Thanks for the summary of the progress and action items.

[pvelissariou1]

Thank you Saeed. This Github discussion is very useful as we track issues/action items/progress very nicely. @ufuk Turuncoglu ***@***.***> Ufuk, could you please add a brief summary (item) regarding ADCIRC and the issues you are facing (e.g., negative element ids) and the steps you are taking to address those? Panagiotis Velissariou, Ph.D., P.E. UCAR Scientist National Ocean and Atmospheric Administration National Ocean Service Office of Coast Survey CSDL/CMMB Physical Scientist - Project Lead cell: (205) 227-9141 email: ***@***.***

…

On Tue, May 30, 2023 at 1:00 PM Saeed Moghimi ***@***.***> wrote: @pvelissariou1 <https://github.com/pvelissariou1> Thanks for the summary of the progress and action items. — Reply to this email directly, view it on GitHub <#5 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/APC7TP7P3SSWAHJTA3BFKULXIYYUXANCNFSM6AAAAAAYUGJL3E> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[uturuncoglu]

@saeed-moghimi-noaa @pvelissariou1 I think I solved the issue with ADCIRC but I still need to verify. I am currently implementing a debug option in there that will give me capability to write import/export fields in VTK format. So, I could visualize the results. Once I have more information I'll share with you under ADCIRC issue page.

[pvelissariou1]

Meeting 02/26/2024
Participants: Panagiotis Velissariou, Ann Tsay, Jana Haddad, Ufuk Turuncoglu, Christopher Domanti, Saeed Moghimi, Yunfang Sun, Keven Blackman

Agenda:

1. UFS-CAT tutorial slides
   - Comments & additions
   - Finalize the slide deck (by 02/28/2024)
   - Presenter(s) ?
2. Model component priorities in UFS-Coastal
   - SCHISM, ROMS, ADCIRC, FVCOM ?
3. Sandy tests for Shinnecock

[janahaddad]

UFS-Coastal Monday tag-up March 18 2024

Agenda:

• README & License files for UFS-Coastal Update ReadMe of ufs-coastal #47

• current iteration priorities: now to April 1 https://github.com/oceanmodeling/ufs-coastal/milestones

  • NCAR/Ufuk focus:
    • SCHISM & ROMS issues and RTs
  • SSM Team (Takis, Yunfang) focus:
    • WW3 with BCs - Yunfang WW3 boundary condition issue WW3#5
    • SCHISM, ADCIRC, WW3 RTs https://github.com/oceanmodeling/schism/issues/3 - Takis
      • when tested & passing, will add to Wiki & need to add to ufs-coastal

[janahaddad]

UFS-Coastal Monday tag-up April 1 2024

If possible, discussion updates should be recorded here

Agenda:

• Ufuk updates
  • Discussion with UW folks
  • SCHISM/ADCIRC/ROMS issues
    • Priority: SCHISM, ADCIRC, ROMS
      • For ROMS we need a few simple test cases (we may pickup a few from https://github.com/myroms/roms_test)
• Discuss storage and accessibility of RT data files and configs (cloud?)
• Update on readme file, Update ReadMe of ufs-coastal #47
• Update on new SCHISM (and other) configs oceanmodeling/schism#3
• Update from Yunfang on WW3 with Open BCs
• Prep for Iteration Planning Meeting if needed

---

Meeting highlights:

workflow brainstorming

• Ufuk had discussions with UW folks at their regular meeting last week. Short Range Weather App is going to be using Unified Workflow and is beginning to define what they need, how they will collaborate with UW team.
  • Might be worth a conversation with SR-Weather folks to hear how they plan on contributing and try to join forces
  • UFS-Coastal team should review https://github.com/ufs-community/uwtools/ before our Milestone Planning Meeting
• the way we currently see it, the UFS-Coastal Workflow should include tools that allow a user to pre-process model inputs, set up the model components as needed, deploy a coupling application (compile & run), and post-process coupled model outputs.

SCHISM/ADCIRC/ROMS open issues

• Specified UNIT in FLUSH is not connected with GNU roms#2 -- need support from roms devs on this to move forward
• may need to fork myroms/roms_test repo into oceanmodeling at some point
• currently have DATM+ROMS application RT in our list
• Ufuk recently updated SCHISM version
• Takis' ADCIRC pull request accepted (??) to address adcirc coupled with ww3 is not working adcirc#1
• Discuss storage and accessibility of RT data files and configs (cloud?) [@pvelissariou1 can you add your notes on this item if any?]

draft ReadMe file

• UFS-Coastal team will review and comment on Takis' draft readme
• remove duplicated items in the Sphinx docs

list of updated SCHISM, ADCIRC, WW3 RTs in #58

• Takis is currently testing these on Orion
• Will next test concurrently on Frontera and Hercules, Hera

WW3 standalone with open BCs

• Yunfang is seeing some open boundary nodes are showing up inside the model domain, will bring up with WW3 devs this afternoon

Milestone Planning Meeting will include:

• discussion about what the initial workflow should include and how this relates to. Will try to define requirements/requests from leadership prior to that. (@dpsnowden @saeed-moghimi-noaa)
• how this relates to ESMX

@uturuncoglu @anntsay @yunfangsun @pvelissariou1 please review if I missed something!

[janahaddad]

thanks @pvelissariou1 !

[janahaddad]

UFS-Coastal Monday tag-up April 5 2024

Agenda:

Jong Kim from EPIC

• ufs-coastal structure
• EPIC's Jenkins workflows
• EPIC's work on updating/simplifying rt.sh

Review current iteration tasks

[saeed-moghimi-noaa]

Jong's slides: https://docs.google.com/presentation/d/1adoCP6-mueo-yXxlTRv6SyysbAHWZivQDJ6atYB7QEU/edit#slide=id.g2c87049156c_0_0

[janahaddad]

UFS-Coastal Monday tag-up May 13 2024

Agenda:

• Update from Ufuk re: UW Tools testing
• UIFCW 2024 submissions
• reminder to provide feedback on UFS-CAT Round 3 planning
• other team-member updates & and dashboard review

[janahaddad]

Meeting notes

• UW Tools testing: possible approaches to ingest workflow in application layer ufs-community/uwtools#478 & Testing UW tools #78

  • currently no support for ESMF config (required by CDEPS) format, only support for YAML. Open to expanding, and we have a workaround
  • UW reads config.yaml where all implementation configs are set
  • in person meeting this week re: coupling support for UW
  • Ufuk's aim is to reproduce DATM+SCHISM configuration
    • short term is to create set up nems file
    • next step is to implement Jinja Template

• Workflow exploration: Ufuk suggests holding a technical meeting to share findings (Ufuk presents on UW, Armaghan presents on STOFS Workflow & others)

• pySCHISM

  • can someone on SSM team use pySCHISM to replicate inputs for DATM+SCHISM configs ?
  • Jana will put Ufuk in touch with Soroosh and/or Bahram & will share some documentation on pySCHISM
    • Ufuk likely won't have time for this before his trip. Will revisit at that time.

• Meshes:

  • per Takis there is a repo with a set of meshes (STOFS repo? @gseroka may be able to help locate)

• UIFCW abstracts:

  • Ufuk will drop his 2023 abstract/talk in the GDrive
  • Ufuk & Ann will connect with Christina @ UW tools coupling brainstorm in Boulder this week to align presentations

• Ufuk met with Hernan re: GNU compiler issue

  • they are compiling with their home script and we're compiling with ufs-coastal
  • Ufuk sent his run directory to them to test, and they're testing on their side
  • this issue is not critical

[gseroka]

Meshes:

• there is a repo with a set of meshes (STOFS repo? @gseroka may be able to help locate)

This may be the repo you're talking about:
https://github.com/noaa-ocs-modeling/STOFS-operational/tree/main

We didn't upload the meshes to there due to GitHub file size constraints. We do have meshes on GDrive and AWS for STOFS-2D-Global and STOFS-3D-Atlantic if needed.

[janahaddad]

ok thanks @gseroka, this is useful to keep in mind for workflow planning down the line

[janahaddad]

UFS-Coastal Monday tag-up May 20 2024

Agenda

• Update from @yunfangsun on WW3 (no obc's) + SCHISM coupling
• Review lingering open items, and updates from Ufuk on UW coupling meeting
• UIFCW abstract submissions
• Discuss moving landing page to ufs-coastal-app and necessary steps (e.g. rename ufs-caostal to ufs-weather-model as fork)
• Jana update on EPIC S3 bucket for input file storage
  • is this a place for storing RTs as well?
• Take-aways from UFS-WM model infr dev mtg today
• Preparations for Ufuk's leave period

[janahaddad]

Meeting notes

• Take-aways from UFS-WM dev meeting today:

  • Ufuk shared about testing he's doing with UW tools. He has implemented creation of CDEPS namelist files with UW.
  • UW Coupling brainstorm: Ufuk supporting/advising on design of capabilities for coupled in UWtools.
  • UFS-Coastal is working on supporting Frontera for testers/evaluators. There's now a PR on UFS side to add module file & build sys support. UFS-Coastal is having some issues with spack-stack install on Frontera.

• UW Coupling brainstorm: take aways from Ufuk & Ann

  • CDEPS implementation is almost done and seems to working ok
  • we can create the DATM+ X workflows
  • UFS-Coastal will explore/try other workflows in parallel
  • mention desire to use UW tools at UIFCW talks

• Frontera support by UFS-WM: Feature/detect frontera ufs-community/ufs-weather-model#2265

• RT system:

  • should stay lightweight, want the RT system to be run-able regularly with fast results
  • can have heavier tests on app-level outside of RT system
  • https://github.com/myroms/roms_test/tree/main/IRENE/Coupling/roms_data_cdeps
  • Yunfang is working on this and will start with building the test cases for SCHISM only and DATM+SCHISM on Shinnecock mesh after H. Ian cases listed/planned in SCHISM+WW3 H. Ian testing-suite: Phase 1 #92

• 2D to 3D SCHISM-WW3 coupling prep

  • set up quick mtg with Joseph Denise W. to determine units etc. Also add EMC folks

• move to ufs-coastal-app

  • targeting Wed 5/28 for the move (see steps in model repo name change & transition to coastal-app repo as landing page #71)

[janahaddad]

UFS-Coastal Monday tag-up June 3 2024

Agenda

• Update from @yunfangsun on WW3 + SCHISM Phase 1 coupling for H. Ian case SCHISM+WW3 H. Ian testing-suite: Phase 1 #92
• Updates from @Armaghan-NOAA on workflow exploration connections to UFS Coastal ?
• Review model repo name change & transition to coastal-app repo as landing page #71 with team
• Take-aways from UFS-WM model infr dev mtg today

[janahaddad]

Meeting notes

• Take-aways from UFS-WM dev meeting today:
  • Upgrade spack-stack from version 1.5.1 to 1.6.0 ufs-community/ufs-weather-model#2093
  • SFS (? not sure what this is) is considering adding WW3 coupling with CICE. Question that came up is would WW3 be added on the "inner" or "outer" coupling loop. Denise says has to be inner to match coupling time step. They also are leaning toward using reg. rather than unstruc. WW3 grid. Add regression testing for wave-sea ice coupling ufs-community/ufs-weather-model#2072

[saeed-moghimi-noaa]

Thanks @janahaddad

[janahaddad]

Had a tag-up Friday June 7 with @pvelissariou1 @yunfangsun @Armaghan-NOAA @saeed-moghimi-noaa about how to handle PR #68, recent changes in cmmb branch, and related topics. Here's the outcome of our meeting:

• Let's scrap Updates on the test cases and minor fixes in the rt* scripts #68 since several decisions have been made over the last couple months that make much of this PR "unplanned" now.
• @pvelissariou1 will open a new PR only for the recent SCHISM updates that allow complete standalone build with Utilities. This has now been merged into the SCHISM-Dev/SCHISM main branch.
• https://github.com/oceanmodeling/ufs-weather-model/tree/cmmb currently points to Takis' cmmb branch on SCHISM-Dev -- Takis will change this to point to the commit hash that coincides with the current SCHISM-dev/SCHISM main.
  • Once that's done, the next step is to share the oceanmodeling/ufs-weather-model/tree/cmmb commit hash with the team here: coastal_ian_hsofs_schism #100 so that Yunfang can run some lightweight tests on the schism standalone CSTLS build
    • @yunfangsun can you create an issue here when you start this step?
• For the oceanmodeling/ufs-weather-model/tree/cmmb branch: This should ideally be renamed to something more descriptive of what it is right now -- feature/schism-stdalone-update ?? When @uturuncoglu returns and merges this that branch can then be deleted.
• At some point soon, as user base expands, the ufs-coastal team should develop/agree on a future git workflow:
  • branch naming, branch lifetime,
  • PR structuring,
  • cadence of syncing with upstream model components (should prob be model-specific)
  • how to fix a working version for users -- using tags & releases at the app-level ??

Thx all!!

[janahaddad]

UFS Coastal Monday tag-up June 10 2024

• Reviewing Yunfang's updates on SCHISM+WW3 H. Ian testing-suite: Phase 1 #92
  • will use coarse SCHISM mesh for the official "H. Ian testing suite" rather than HSOFS or Subsetted Atl mesh
  • post-processing with Ali's wave-tools: discuss with @aliabdolali tomorrow Tues 6/11
  • @yunfangsun ran global model (plz add update to Run global UnWW3 to provide boundary to Atlantic set up WW3#9)
  • RT system SCHISM component updates: put this step on hold for now and revist when SCHISM+WW3 H. Ian testing-suite: Phase 1 #92 is complete. Note that these RT system updates are being tracked in this master issue SCHISM and WW3 RT system cases (main issue) #95
• Reviewing @Armaghan-NOAA's Extraction of Workflow demo:
  Flowchart_python_v04_Armaghan.pptx
  • take-away: was a good thing to test out, and helped sort the various STOFS scripts into 1) pre-processing, 2) nowcast/forecast, 3) post-processing -- but it's not a great tool for comprehensive flowchart as is. Prob not worth too much more effort. @Armaghan-NOAA if you want to have these python scripts around for potential updating/collaboration in the future you could create a personal repo for them.

[janahaddad]

UFS Coastal WW3 tag-up notes Monday June 10

attendees:
@platipodium @saeed-moghimi-noaa @yunfangsun @sbanihash @janahaddad @Armaghan-NOAA

topics discussed:

• Status of Improve WW3 mesh cap to allow using vortex formulation in ocean side WW3#11 and general discussion of how to best move this forward
  • Ideal/optimal solution is to have variables needed calculated in mediator layer rather than in cap, but this solution is in this case largely limited by the size of spectral data (38 freq. bins x 32 directional bins per x,y)
  • @platipodium comment:
    • WW3 folks: dump your routine here, and let @josephzhang8 Joseph/me implement it properly https://github.com/schism-dev/schism-esmf/blob/master/src/schism/schism_esmf_util.F90
• limited resources on EMC WW3 team curerntly to support this and future needs. Saeed is working on procuring resources
• EMC reminder to regularly sync with the UFS-WM version of WW3 (each 2-3 months at most). I believe this comes with each UFS-WM upstream sync that @uturuncoglu does, which is typically done monthly. Will review model component update cadence when Ufuk returns from AL.

Thx all! please add a reply if I missed something important

[janahaddad]

UFS Coastal tag-up Monday Sep 23 2024

• Joey Smith ROMS CICE coupling
• Worklow updates
• WW3 + SCHISM coupling issue
• RT system updates

if there is time:

• FY2025 priorities review
• roadmap advertizing
• Jenkins CI/CD

[janahaddad]

Meeting notes

Joey Smith ROMS+CICE coupling

• Implementation of ROMS+CICE coupling roms#4 & ROMS-CICE coupling for Bering Sea roms#5
• https://github.com/myroms/roms/blob/develop/Master/cmeps_roms.h
• Ufuk suggestion is to compare the existing CICE cap with the one in UFS Coastal (WM)
  • File called import-export in https://github.com/NOAA-EMC/CICE/tree/emc/develop/cicecore/drivers/nuopc/cmeps

[janahaddad]

UFS Coastal tag-up Monday Sep 30 2024

• ESMF mesh issue
• elev.th issue -- Mansur demo
• Yunfang working on Wiki page: https://github.com/oceanmodeling/ufs-weather-model/wiki/Develop-new-applications-for-UFS%E2%80%90WW3
• Duck FRF observations for Dorian Duck FRF observation information #133

[janahaddad]

Meeting notes

• ESMF mesh issue: Ufuk will follow up
• elev.th issue: as described here: ufs-coastal DATM+SCHISM DuckNC #127 (comment)
  • Mansur will use Type 4 BC (space & time varying) and generate from HYCOM output
  • track issue in oceanmodeling/schism
• Yunfang wiki page plan:
  • UFS WW3 standalone
  • UFS SCHISM standalone
  • UFS SCHISM+WW3+DATM
• Joey updates:
  • working on 2 CDEPS caps, one for atm data and one for ocn data
  • working on purely atm configuration as a debugging step
  • next: create a configuration for datm standalone model
    • @SmithJos13 a couple potential resources for you: UFS-CAT Troubleshooting help #46, UFS-Coastal Walkthrough & Office Hours #93, SCHISM+WW3 H. Ian testing-suite: Phase 1 #92 (all the completed issues here contain a path to the configuration in Yunfang's Hercules directories)
• Workflow updates
  • Mansur working on input generation script for schism, will update and share with Ufuk here SCHISM pre/post processing integration in workflow  ufs-coastal-app#9
  • Ufuk aims to have workflow capable of running datm+schism in a couple weeks or so

[janahaddad]

UFS Coastal WW3 tag-up

• Improve WW3 mesh cap to allow using vortex formulation in ocean side WW3#11 (comment) wave number not exchanging
• exchange of SSH in schism ww3 coupling: DATM+SCHISM+WW3 water level exchange (coarse resolution SCHISM mesh) #132

[janahaddad]

UFS Coastal tag-up agenda Monday Oct 7 2024

• Joey updates
• Jenkins: Ufuk needs to change rt.sh, it’s otherwise all set
• RT system Duck cases
• elev.th file issue –
• esmf mesh issue
• Stakeholder meeting kick off plan

[janahaddad]

meeting notes

• Jenkins/automated testing Automated reg-testing  #42 :
  • changes were needed in rt.sh to correct paths to RT inputs and new directory for outputs
  • will create new base directory in work2/nosofs ...
• RT system Duck cases SCHISM and WW3 RT system cases (main issue) #95

[janahaddad]

UFS Coastal tag-up agenda Monday Oct 21 2024

• Stakeholder meeting retrospective
• BIL annual mtg updates
• coastal_ian_atlantic_datm2sch2ww3 #124 (comment)
• Improve WW3 mesh cap to allow using vortex formulation in ocean side WW3#11

[janahaddad]

Meeting notes

• Stakeholder meeting
  • we'll continue on monthly frequency
  • action items to follow up:
    • google form for feedback
    • sharing recording and links shared during the meeting
    • follow up on CO-OPS request for focused convo
  • backlog items/ideas to discuss/prioritize from the meeting:
    • adding WW3 support through UWtools (UWtools team has a starting point WW3 driver)
    • adding active atm through FV3... if collaborator/interested-party is willing to work on
    • next sync with UFS-WM (@uturuncoglu didn't get a chance to ask today, what timeline are you thinking?)
    • debug version o RT's to add to rt_coastal.conf?
• Ilya Rivin joined from Co-ops modeling team.
  • curious about one way coupling with ATM
  • NWM coupling is biggest interest
• coastal_ian_atlantic_datm2sch2ww3 #124
  • re: interpolation issue, @uturuncoglu got help from ESMF team and they tracked the issue using debug version of ESMF. Bob at ESMF is working on a fix, then @yunfangsun can test Ian case again
  • if interpolation fix tests are positive then we'll move forward despite the issue related to WW3 scrip generation (causing mismatched meshes between WW3 and SCHISM). The cause is internal to WW3 code (and cap), so we'll also reach out to WW3 folks to share the issue and have their thoughts
  • moving forward as-is will mean always using different meshes for schism and ww3 in a coupled config. i.e., even if feeding in identical grids, ww3 will create it's own. It's not fully clear at this point exactly what ww3 is doing to the mesh, but it seems to add addiitonal nodes.
• Improve WW3 mesh cap to allow using vortex formulation in ocean side WW3#11
  • which mesh to use moving forward for this Duck case?
    • use Ali's WW3 mesh & ask Dan to create SCHISM set up
    • @yunfangsun will reach out to Dan as needed
• @mansurjisan working on ufs-coastal DATM+SCHISM DuckNC #127 (comment)
  • & exploring UW tools API set up at app-level
• @SmithJos13 troubleshooting help for DOCN_CICE case

[uturuncoglu]

@janahaddad @yunfangsun @mansurjisan @pvelissariou1 @saeed-moghimi-noaa JFYI, i have just synced coastal fork with UFS WM. Since, we have extra variables in the mediator history files, I need to create new baseline for the all configurations in rt_coastal.conf. These answer changes are not related with the model components and model components output just bit-to-bit with the previous baseline. Also, note that I did not sync the coastal specific components yet. That would be the next thing. So, I am going to step by step here since if we have issue in the RTs, it would be hard to find the source of the issue (due to the UFS WM sync or updated coastal specific component).

[uturuncoglu]

I think we also need to create new baseline also on Frontera. So, you could see RT failures on the Frontera and Hercules due to the baseline changes at this point. Once those are updated, it would be good to go.

[janahaddad]

Thanks @uturuncoglu

[uturuncoglu]

@janahaddad @yunfangsun @mansurjisan @pvelissariou1 @saeed-moghimi-noaa @danishyo I also created new baseline on Hercules and run all RTs against it. It all good now except FVCOM. We know that it has reproducibility issue but not in our priority list at this point. There was a recent UFS WM PR related to Frontera support (ufs-community#2265, combined and merged with other PR). The next thing is to run the synced model on Frontera to see what happens and also create baseline. But, before that I am plaining to update coastal specific components one by one and test them.

[josephzhang8]

Add @danishyo

[janahaddad]

UFS Coastal tag-up agenda Monday Oct 28 2024

• EPIC has drafted a UFS Coastal success story for posting on UFS website
• Follow-up to prepare for NOAA AWS S3 data bucket: ufs-coastal baseline maintenance and storage #84 -- file structuring the existing cases for copying to S3
• update on syncing with ufs-wm and authoritative model repos (see below)
  • update SCHISM (tested standalone & in PSurge simulations)
  • update PaHM (tested standalone)
• new SCHISM RT cases (in progress, need to update schism's config files)
• ufs-nemspy
• Duck DATM+SCHISM case
• Joey's docn+cice case

[janahaddad]

UFS Coastal tag-up agenda Monday Nov 4 2024

• Questions from Takis
• Joey troubleshooting #Implementation of ROMS+CICE coupling roms#4
• Ufuk meeting debrief re: FVCOM for UFS CAT troubleshooting
• status of ESMF Library debugging for the mesh issue?
• ufs-coastal DATM+SCHISM DuckNC #127 (comment)
• https://earthsystemmodeling.org/esmpy_doc/release/latest/html/install.html & https://xesmf.readthedocs.io/en/stable/

[janahaddad]

meeting notes

• @pvelissariou1 setting up 33 diff storms for SCHISM+PAHM using Atl schism domain
  • 1-2 of those cases could go into the RT suite, Shinnecock domain for H. Sandy (pahm) (internal cap)
  • Ufuk asks to create a PR for those RTs, pointing to the data paths
• Joey and Ufuk will set a time to meet re Implementation of ROMS+CICE coupling roms#4
• FVCOM cap only accepts pressure & u,v -- Sebin will make changes suggested by Ufuk and try again
• @uturuncoglu sync'd with ufs-wm, and then upgraded model components one by one
  • ADCIRC failing -- Ufuk will test with latest release 56.02
  • ROMS failing -- Ufuk talked to Hernan who will try to update
  • FVCOM -- did not update
• ufs-coastal DATM+SCHISM DuckNC #127 -- @mansurjisan to try instead generating uniform wind field from FRF obs over the domain

[janahaddad]

UFS Coastal tag-up agenda & notes Monday Nov 18 2024

• Team goal: try to be a bit more descriptive when creating new issues. Try to include an overall description of the goal/objective, or the problem to be solved.
• Joey's case is working!
  • Worked with Ufuk, they decided that they need to use mediator
  • Introducing mediator resolved many of the issues
  • Making sure data is being shared on CICE side is important for correctly setting BCs, still working on that
  • It’s also running really slowly right now, so may need some optimization
    • Ufuk suggested ways to get some speedup: ESMF_profile_summary is created, and that help you understand timing; coupling interval every 60 sec may be too much;
    • export ESMF_RUNTIME_PROFILE=ON
    • export ESMF_RUNTIME_PROFILE_OUTPUT="SUMMARY"
    • … these need to be added to your job script to create the ESMF_Profile.summary
  • Questions about coupling timesteps
    • Need to modify run sequence and to match with the time steps within the models themselves
  • Next step is to validate this CICE configuration, then can ingest to RT system
• EMC WW3 team is now reviewing PRs again
  • Add fields and fixes to dev/ufs-weather-model branch, as required by Coastal App NOAA-EMC/WW3#1291 (comment)
  • Modifications for ufs-coastal app ufs-community/ufs-weather-model#2396.
  • Once these changes are reviewed/merged at EMC/WW3 and UFS-WM, Ufuk will sync again
  • Also looks like EMC is getting ready to add ability to produce WW3 history files and restart files with PIO.
• WW3 gridding issue
  • coastal_ian_atlantic_datm2sch2ww3 #124 (comment)
  • WW3 identical mesh issue in coupled applications NOAA-EMC/WW3#1319
  • WW3 devs agree this needs to be addressed but for now this means
• RT inputs uniform across configurations
• Yunfang DATM+WW3+SCHISM case for Sandy Duck:
  • Comparisons are using different switch files ...
  • Working on
  • Duck_coupled slide 10
    • all but "Dan WWM" are standalone, all are outside UFS Coastal
    • working on brining into ufs coastal now
• Mansur DATM+SCHISM case for Sandy Duck: ufs-coastal DATM+SCHISM DuckNC #127
  • have the gr3 file provided by Dan (he re-formatted from the WW3 mesh provided by Ali)
• Recent update of ADCIRC model is causing issues … Issue with updating ADCIRC adcirc#4
  • We agreed to contineu with the version of ADCIRC (v55.xx) that works in UFS Coastal. When STOFS-2D-global updates to v56, we could reconsider updating at that time; will require re-generating the atm_adc hotstart nc files in those RT cases (or fully updating the RT case to Duck domain?).
  • This highlights the need for good documenation and metadata of our RT cases! All new RT contributions to UFS Coastal should come with metadata/description
  • General approach for model updates sould be to point to commit that is stable within UFS-Coastal unless we have some sort of direct support from model developers or contributors for that model
  • Ufuk noted that there are some contributions into ADCIRC we will need to confirm are present in any version we use (e.g. ghost element handling)
  • @saeed-moghimi-noaa @uturuncoglu please comment here if needed, want to make sure we're all on the same page moving forward
• Helping ufs-cat troubleshooting … next week is stakeholder meeting
• status of ESMF Library debugging for the mesh issue
  • Bob at NCAR has updated ESMF library with fix; Ufuk is testing with Yunfang's high res case, but it appears to be hanging, need to look at more carefully, will continue to update.

[pvelissariou1]

Thursday 12/05/2024 UFS-Coastal Presentation at NHC (1:00 - 2:00 ET)

• Presenter P. Velissariou
  • starting with an updated UFS-CAT tutorial and a Hurricane Beryl 2024 test case
  • invitations will be extended to this group

[janahaddad]

good to know! thanks for sharing @pvelissariou1, will you primarily focus this on SCHISM + PaHM coupling?

[pvelissariou1]

Yes, primarily, SCHISM+PaHM (internal and through CDEPS) at this point, for demonstration purposes.

[janahaddad]

UFS Coastal tag-up agenda & notes Monday Dec 2 2024

• Issue with updating ADCIRC adcirc#4 (comment)
  • will stay with 'working' version (v55.xx) & then stay with STOFS-2D-G version from here on out
• ESMF mesh issue update ?
  • see item 1 here for background: coastal_ian_atlantic_datm2sch2ww3 #124 (comment)
  • arising for the edge case with sch-ww3-datm case with H. Ian...
  • testing on NCAR machine for now, with Bob O. @ NCAR
• schism-3d branch not working with latest sync:
  • may be related to UFS-WM updates needed... creating an issue on Frontera, for which Ufuk did a hacky workaround
  • Ufuk will test & brought it up at UFS-WM dev meeting today; relates to the spack-stack & ESMF library versions
    • related to Issue with Frontera spack-stack 1.6.0 JCSDA/spack-stack#1388 ... this relates to a higher-level dicussion needed at UFS-WM on how these spack-stack updates are done & the dependency of libraries
  • affect's Dan's testing of VF coupling: Improve WW3 mesh cap to allow using vortex formulation in ocean side WW3#11 (comment)
    • @danishyo: Ufuk plans on providing a quick fix to this modules-related problem soon
• @yunfangsun : WW3 Homogeneous case (H) option (not woring with UFS Coastal)
  • ufs-coastal DATM+WW3 DuckNC #128: a couple issues related to UFS-WW3 standalone & DATM+WW3 to bring up at WW3 tag-up later today
• Joey: load balancing ...
• @mansurjisan : working on running Duck case for DATM+SCHISM with @aliabdolali 's Sandy set-up
  • use oceanmodeling/ufs-weather-model default branch (will not have cap updates for WL exchange or 3D VF coupling )
• Takis giving a talk to NHC on SCHISM+PaHM (this Thurs)
  • @pvelissariou1 please share your draft slides?
  • Joseph Zhang has incorporated PaHM fixes in SCHISM; Takis has tested successfully;
  • Ufuk needs to sync both SCHISM and PaHM

[janahaddad]

UFS Coastal tag-up agenda & notes Monday Dec 16 2024

• Yunfang:
  • ufs-coastal DATM+WW3 DuckNC #128
  • schism3d branch water level forcing
• Mansur:
  • ufs-coastal DATM+SCHISM DuckNC #127 (comment)
• Ufuk:
  • ESMF 8.8 will be released in January, but won't inlcude anything for UFS-Coastal
  • issue rel. to interpolation is not yet fixed in ESMF; still in debuggining; Bob will start working on it after the next ESMF 8.8 release
  • next release will be in the spring; beta in the interim will hopefully have a fix
  • Saeed: can we invite Bob to engage with the team
  • UW Tools/app-level dev:
    • data retrieval and ESMF grid gen is working ok
    • meeting with UWtools: need CDEPS configurabiloity to config.yml
• CO-OPS team (AJ, Xu, Lucila)
  • AJ: testing on WCOSS2
    • Mansur has access to WCOSS2/Acorn testing env for now
    • Ufuk: note that UWtools needs to be installed with spack-stack
• Joey: will create a new RT via PR for cice-datm-docn for a small lighweight case
  • Ufuk: Create a PR for the CDEPS, bring sub-PRs for the models, Ufuk will help
  • Joey working on a new clean fork to incorporate; can set a separate "how to PR" meeting that we'll record for future contributors

[mansurjisan]

I am working on getting access to the WCOSS2 Testing and Development System (TDS), Acorn cluster. However, I do have access to the WCOSS2 Development cluster, Dogwood.

[pvelissariou1]

@mansurjisan shouldn't you have access to all dev wcoss2 clusters by default? I haven't checked it in detail from side.

[mansurjisan]

@pvelissariou1 ,

no, I actually requested access specifically for the development cluster (dogwood) where I am experimenting with the workflow associated with existing operational ocean forecast systems as well as planning to run some UFS-Coastal related regression tests. I need to submit a separate request for accessing the TDS system (Acorn).

[pvelissariou1]

@mansurjisan I see, I'll check what clusters I have access to. For sure we don't have access to the production ones. Thanks.

[janahaddad]

UFS Coastal tag-up agenda & notes Monday Jan 6 2025

• Ufuk:

  • will get path to STOFS-3d-atl directories on Hercules (copied from WCOSS2 runs)

• Mansur updates/overview of progress on workflow-level scripts

• Yufang updates/overview of progress on WW3 case for Duck, NC domain

  • need to submit PR for updates to CDEPS and CMEPS needed for this configuration

• Soroosh issues with PaHM cap & how to write out full ESMF output?

• Xu: questions on running SCHISM standalone on WCOSS2

• Joey: comparing CICE inside (docn datm cice) and outside UFS Coastal & seeing differences

• S3 bucket on NODD -- need to share RT directory with EPIC folks when ready

• introducing new regression tests for Duck domain:

  • we'll start with DATM+SCHISM config
  • discuss next week
  • record a "how to create/contribute a new RT"
    • include some metadata, descriptions, maybe a few figures to document
    • structure in a heirarchical way