CMWG: Meeting notes 2023 (Archived)

Cryosphere Working Group
6

Cryosphere Modelling WG meeting:
13/Dec/23
Attendees: 13

Merit allocations:

Update from co-chairs:

  • Chen has joined as a co-chair

Review responses from developers (initial contact)

  • Spreadsheet: Ice sheet models - Google Sheets
  • Notes column shows initial feedback on whether there is capacity to support with coupling into ACCESS
  • Olga Sergienko is developing an ISM and will be part of MOM6 but at this stage there is no documentation on what that models has; although SSA model (no higher-order physics) but she’s happy to answer questions in the Developer Questionnaire.
  • Overall, happy with the initial feedback.
  • Comment from group about CISM could be easy to follow from the ocean/sea-ice perspective; will need to weigh the ease of coupling vs other aspects like the physics, etc.

Review google form responses to ToRs doc

  • 13 responses (6 ice sheet modellers)
  • Only 2 questions where we reached consensus
    • Hydrofrastrucre/surface hydrology scheme available (nice to have)
    • Code management for community development (essential)
  • Broader agreement across:
    • Code management for community development (essential)
    • Higher-order physics (essential)
    • Open and accessible code (essential)
    • Mass/energy conserving (essential)
    • Relationship with developers (essential)
    • Unstructured mesh (essential)
    • Automatic differentiation (nice to have)
  • Remaining criteria was less clear
  • Felicity will be doing a full summary and will share
  • Suggested next steps:
    • We had 10 criteria (including physics/schemes) that were identified as essential
    • Pending review, adopt these essential criteria
    • Determine which models adhere to these (including through discussions with developers)
    • Exclude models that do not meet these criteria
  • Relationship with developers generally considered essential
    • What minimum requirements do we expect from developers?
    • Do we exclude any models now if developers have said they are unable to support? (Grisli, KORI, MALI)
  • Higher-order physics generally considered essential
    • Do we exclude any models that do not adhere?
  • Discussion:
    • Developer support
      • We want really good relationships, open, sees advantage that Australia can provide
      • Want a 2-way relationship, and a way to get support
      • Open to making changes
      • Established user community, having more than just a single developer and a community to engage and tap into for expertise.
    • Suggestions
      • Could aim to have the nice to have category of criteria be things we might want to see down the road.
      • Rule out those that can’t offer development support now.
      • Wait on ruling out based on higher-order physics to chat with developers (check before eliminating)
      • Could also recast the developer relationship to how welcoming they are to suggestions for variations / modifications.
  • Request to the WG - need help populate spreadsheet on the potential models: Ice sheet models - Google Sheets
    • Timing: aiming for early January
    • Volunteers from meeting:
      • Roland - SICOPOLIS
      • Poul - CISM
    • Start with essential criteria - if we find out any lacking, then we don’t worry about fully populating

Timeframe for determining ISM

  • Publish essential/desirable criteria (aiming for end of Dec 2023)
  • Developer questionnaire (early Jan)
  • Aiming for March for the expert panel

Other things

  • Ethics subgroup: Pat, Lawrence, Kelsey
    • Currently adopted COSIMA code of ethics in the interim but will this group will work on feedback from September WG meeting and if we want to develop anything further
    • Compute subgroup: Dan Atwater, …any other volunteers?
  • Will share a spreadsheet calling for any reference datasets that the WG would like to see available at NCI?
  • Next year we’ll be going to monthly meetings - please remove the 2023 calendar invite if you have it locally added.
  • Next meeting will be in March - we will add the scheduled dates/info to the Hive.