CMIP7 evaluation group meeting notes

CMIP7 CMIP7 evaluation
1

Some notes from our first meeting - please feel free to add or edit as necessary.

Topic of discussion: What is the scope for the upcoming evaluation hackathon (to be held as a training activity during the September ACCESS workshop)

  • There seems to be three broad categories for evaluation:
    1. Live/‘on the fly’ diagnostics, more relevant to core development team
    2. ESMValTool for more generalised metrics, more of a community effort
    3. ILAMB - more used by the land surface modellers

Hackathon focus on ESMValTool might be more beneficial for the wider community.

  • What kind of recipes to focus on?

    Suggestion was made to develop beforehand a generalised recipe which can be modified to read in variable of choice and plot some standard maps/metrics. Eyring et al. (2015) has a comprehensive list of recipes - perhaps just selecting a few and working on customising these would work well for a 1-day activity.
    eyring.esmvaltool.gmd16.pdf (ucar.edu)
    Reproducing Figures 2 & 4 might be a good starting point.
    Next meeting 21 June.

1 Like
2

Meeting notes 21/6/24

Romain: Update - progress is being made on setting up recipes as per last meetings suggestions, chasing down ERA5 issues on NCI.

Hackathon: Proposed 2 sessions during the training day. Morning - getting the general recipes to run, afternoon - customising the recipes.

During the workshop it would be good to have a breakout session with general discussion to follow up on ideas from the hackathon. Followed by an online wrap-up the next week to consolidate thoughts and come up with a workflow/plan.

Harun’s suggestion: there is currently a gap between model evaluators and developers. Would be good to begin to reach out to model users to get their feedback - will follow up on this.

Next meeting 12 July.

1 Like