New URL for NEMO forge!   http://forge.nemo-ocean.eu

Since March 2022 along with NEMO 4.2 release, the code development moved to a self-hosted GitLab.
This present forge is now archived and remained online for history.
DevelopersCommittee/Agenda/2020-12-17 – NEMO
wiki:DevelopersCommittee/Agenda/2020-12-17

Version 11 (modified by mikebell, 3 years ago) (diff)

--

Draft minutes for NEMO Developer's Committee meeting Thurs 17 Dec 14:30 - 17:00 Central European Time

The meeting took place by videoconference on Monday 9 November 2020

Participants:

  • Chairperson: Mike Bell
  • External experts: Joanna Staneva, Fred Dupont, Rachid Benshila, Olivier Aumont
  • Consortium Experts : Momme Buttenschon (CMCC), Helene Hewitt (Met Office), Adrian New (NOC), Julien Le Sommer (CNRS)
  • NEMO Officers: Andrew Coward (NOC), Sébastien Masson (CNRS), , Guillaume Samson (MOI), Amy Young (MetO)
  • Working groups leaders: Italo Epicoco, Jérôme Chanut, Tomas Lovato, Guillaume Samson, Martin Vancopenolle,
  • NEMO Project manager: Claire Lévy
  • Members of NEMO System Team: Laurent Brodeau, Clément Bricaud, Diego Bruciaferri, Stefania Ciliberti, Emanuela Clementi, Nicolas Martin, Francesca Mele, Clément Rousset, Sybille Techene
  • Permanent Invitees and developers with an action in the work plan): Miguel Castrillo, Gurvan Madec, Oriol Tinto
  • Others: Pier Guisepe Fogli,
  • Apologies: Ed Blockley, Dorotea Iovinno, Dan Lea, David Marshall, Jean-Marc Molines, Paolo Oddo

1. Roles and membership of NDC

The proposed revisions and clarifications of the roles and membership of the NDC (see slides) were agreed.

2. Remaining reports from WGLs

There have been a lot of developments to the NEMO code, mainly within IMMERSE, on the 1D Atmospheric Boundary Layer (2 papers); surface waves (a test case is pending); parametrisation of vertical mixing (OSMOSIS); bulk exchange parametrisation; sea-ice coupling and surface current feedbacks.

Guillaume Samson, who is taking over from Seb as WGL, intends to promote more discussion of these developments.

Action: Tomas and Guillaume will discuss the need for coordination between the ASI and TOP groups (for example on light penetration).

Jorn Bruggeman has joined the group (it already includes Olivier Aumont, Christian Ethe, Simon Muller, Renaud Person so is similar in size to the AGRIF WG). Tomas has started to discuss views on TOP with a wider range of BGC modellers.

The work of the group has been quite focused on the tasks identified in the strategy. These plans are now being revisited and analysed in more detail. (particularly wrt vertical sinking and the carbonate system).

The WG has focused on building a coherent roadmap for NEMO V&V. The first draft is now available on the V&V WG wiki page?. This is a useful start but there is a lot of further work to do to improve our current practices. Further work will be included in and is likely to become an important component of future NST WPs (starting in 2021). Recognising this point, Claire Levy has become joint chair of the WG.

  • Processes (Julien) Update from previous meeting in November.

Julie Deshayes has volunteered to help to lead work a) to gather information on how well or not we are representing specific physical processes with NEMO (a matter of sharing existing material) and b) to turn this into discriminant test cases, which could help measure how specific development are improving the situation.

3. Discussion of 2021 Workplan

The draft 2021 Work Plan discussed is available for DeveCom members here.

(Original page is here)

3.1 Review of actions from last year (see slides) (Claire)

The NST work is equivalent to about 10 experts working full time. 2020 was an important year because it was the main year for incorporation of IMMERSE developments.

The changes that have been merged into the trunk over the last year, listed by streams, include work on:

  • KERNEL: quasi eulerian coordinates; HPG scheme; Runge Kutta, do loop macros, SW example
  • AIR-SEA: current feedback; mass flux convection; bulk; ABL-1D; waves
  • AGRIF: vertical interpolation
  • SI3: EAP rheology; Melt ponds; CP preliminary
  • ENHANCE: ocean column properties in NEMO-ICB; ...
  • IOM: restart with XIOS
  • VALIDATION: test cases; Improvements to SETTE; new OASIS test cases
  • HPC: extra halos; initial tiling; mixed precision

Each development includes work on preview, review, validation of merging and documentation. In addition to these developments 116 bug fixes have been made since Dec 2019. Several members of the NDC congratulated the NST on its impressive work this year (see below).

3.2 Comments on WP from WGLs

  • Kernel - Mike
  • supports proposed work; work on penalisation was not anticipated in strategy but is exciting and should be pursued.
  • work on positive definite advection schemes for IMMERSE by Jerome (PPM) is not yet in the plan
  • it might help to identify priorities for work by individuals
  • HPC - Italo :
    • need to prioritise
    • tiling, halo, communication : must be completed as soon as possible
    • HPC actions are not easy
  • PHYPROC - Julien :
    • IMMERSE : validation and pending actions are key.
    • two actions by Jerome on tides.
    • tide potential ok.
    • internal wave drag ok

  • AGRIF - Jerome
    • work on DOMCFG tools should not be neglected
    • z-tilde only currently works for some NEMO versions (e.g. 4.0.1). It has not been brought up to v4.0.3.
  • SEA -ICE - Martin
    • the plan is not 100% in line with the strategy
    • the strategy is due to be brought up to date
    • the documentation is still to be built
    • a list of missing actions has been (or is being ?) established with Ed
    • there is a need to revise the workplan
    • the validation work for IMMERSE needs clarification.
  • TOP - Tomas
    • all items are relevant.

  • OBS - Dan was not present
    • all the actions have been proposed by the Met Office
  • AIR-SEA - Guillaume Samson
    • most developments (other than light penetration) have already been done, the focus should be on testing and validation

  • V&V - Mike
    • a lot of activities not related to the roadmap; mainly regression testing of the new configuration by consortium members (e.g. validation with AMM)
    • the actions for the roadmap are important but not urgent.
  • V&V - Claire
    • scientific results should be a high priority
    • cf CMIP6 validation - not stable…
    • first semester - validation of new version (NEMO 4.2) should be the priority
    • second priority : moving to a new platform, discussion within ST.

3.3 Comments from Consortium experts

  • Met Office - Helene :
    • well done to everybody
    • SI3 major step forward
    • need to prioritise
    • AGRIF - lack resources at the MetOffice?
    • publications on the work done ?
  • CNRS - Rachid
    • most of verification work is planned to proceed as before; not much evidence of evolution

  • CNRS - Julien
    • super impressed by the work. Need to communicate about that.
    • ex : at the upcoming EGU session. in publications ?
    • validation is key, articulation with IMMERSE . but we need clear metrics for success. this is also important on the political side.
    • publications.
    • evolution of the verification system is key. we should not underestimate the time needed for devlopment/adaptation/adoption of new tools / practices.
    • preparation of the new strategy
  • NOC - Adrian
    • concern on sea ice.
    • vertical coordinate should be a priority.
    • AGRIF is important.
  • CMCC - Momme B.
    • congrats to all
    • CMCC has given feedback on AGRIF. Not main stream though.
  • MOI - Guillaume
    • need to check what will be really done on validation. What criteria. Presentation at ST meetings.

Disc with Claire.

Need twin experiments. (MJB doesn't understand what point is being made here)

  • Clement : need to understand why developing SI3 is so slow.

3.4 Comments from External experts

  • Joanna
    • coupling with waves. still room for improvement
    • work with FABM.

3.4 Comments from NEMO Officers and key NST members

  • Andrew Coward
  • caution some work from last year needs tidying up.
  • some of the remaining tasks (eg communications) will require work from many. Implicit work there.

  • Nicolas Martin
    • work has to be done on the collaborative tool + web platform (they use obsolete tools)
    • verification process : trusting.

  • Amy Young
    • points made previously capture her main suggestions
  • Sibylle :
    • RK3 top priority
  • Jerome :
    • where do we stand in terms of performance with respect to other models
  • Italo :
    • adding a periodic monitoring of the performance.
  • Gurvan
    • my priority is on RK3.
    • a lot of “small” thing to work on to tidy this up.
    • In particular :
    • make sure that the new architecture works with explicit free surface (no sub-stepping)
    • then linear free surface case.
    • then whole nonlinear case
    • also shear term
    • thickness weighted time integration.

3.5 Concluding summary (Mike):

This was a useful discussion. There is more work to do to clarify WP priorities and expectations but there seems to be general agreement that high priorities for the first half of 2021 include RK3, tiling and validation of version 4.2. What has been achieved this year by the NST is very impressive. Well done everyone!

4. Connection details

Topic: NEMO Developers Committee Time: Dec 17, 2020 02:30 PM Paris

Join Zoom Meeting https://univ-grenoble-alpes-fr.zoom.us/j/94128441272?pwd=TXR5Z2lJQ2FmSUFFem8xYXFKdWcxQT09

Meeting ID: 941 2844 1272 Passcode: 500932

Attachments (7)

Download all attachments as: .zip