-
Notifications
You must be signed in to change notification settings - Fork 0
April2019Minutes
Cable Minutes - 2nd April 2019 - and this link to corresponding Agenda
1. Tickets list below.
2. Jhan to update committee on the status of the head of trunk.
trunk tagged TRAC link
3. Martin & Gab to update on benchmarking progress.
We now have a working example of an (largely) automatic continuous integration test suite. This involves the user committing changes to the Share/integration branch to trigger the test suite. The code is freely available and can be used by users to test their own changes locally (https://github.com/mdekauwe/CABLE_benchmarking)
Action Martin/Gab: write some text for a Cable Users email update.
4. Martin update on CNP progress.
Martin has created a repository (https://github.com/mdekauwe/CABLE_CNP). This runs single-site(s) CABLE-cnp for flux sites. The code checks out the trunk, builds and runs CABLE-cnp. The code automatically takes care of model spin up and build the necessary met files on the fly. This makes the need for the CABLE site structure redundant.
Action Martin: write some text for a Cable Users email update. Martin to discuss with Vanessa the progress here.
5. Discussion of tag numbers for the ACCESS-ESM1.5 and ACCESS-CM2 versions of CABLE (Rachel to lead, suggestion: ACCESS-ESM1.5 – CABLE2.4; ACCESS-CM2 – CABLE2.5).
ESM1.5 version of CABLE will be tagged as CABLE-2.4-ESM1.5.
Action Jhan: Organize cable.nml namelists to accompany each of these versions.
6. Discussion of mechanism to resolve deadlock: in a scenario where the committee can't agree on a route forward - and to avoid circular discussion - the issue will be referred to Andy and Rachel so they can arbitrate
7. Any other business?
Action Jhan/Martin: draft an update to CABLE users outlining head of the trunk, tags, benchmarking, CNP and Met Office style guide minimum requirements moving forward for new science.
Longer discussion on strategy around ACCESS. When do we do this. AMIP run 20 years? Rising costs. Scheduled testing, every N months? Have you changed a state variable as part of the ticket. Discussion about complexity of UK met office support for state variables.
* fixed indentation (#215)
Martin has checked spacing/indentation.
Action Danny: to check what the Met Office check does
* climate variables incorrectly declared in casa_cable + missing Pdep (#214)
Martin has a branch fixing this.
Action Jhan: committee accepted this change
* Bug fixes to GW scheme (#213)
UNSW has resolved this.
Action Jhan: committee accepted this change
* Adding changes to ALMA convention into CABLE (#212)
Anna and Gab have a fix for this.
Action Jhan: committee accepted this change
* temperature vs potential temperature (#206)
* Some met variables 4D in site runs. Action - someone to add a comment in the code (#200)
Action Gab: said he would perhaps look at this.
* Tiled vegetation cannot be passed through met file (#199); related changes relating to site level simulations, including tiling (#211)
* Allocation wrong in CMIP6-MOSRS when using MODIS phenology (#204)
* partitioning evaporation into evaporation and sublimation (#196)
* inconsistencies in initialization of wbice (#174)
* code clean up in soilsnow module (#169) - who could comment on this? Or close it.
* Calendar attribute for time variable (#159)
Action Jhan/Martin: email Peter Briggs for a link to the fix
* clean up roughness variables (#153)
* problems in C4 simulations (#140)
Action Martin: test this fix with benchmarking suite.