-
Notifications
You must be signed in to change notification settings - Fork 0
March2019Minutes
Cable Minutes - 5th March 2019 - and this link to corresponding Agenda
1. Discuss tickets
2. Update the other committee members on the on steps taken since the previous meeting to fix the working trunk (CMIP6-MOSRS).
Actioned - see Minutes
3. Discuss any issue with now making this the head of the trunk?
Action: Complete
The head of the trunk has merged Share/CMIP6-MOSRS/ and aside from the offline/ driving code is now identical to the ACCESS-CM2 version of CABLE.
This is at trunk revision 5717.
A corresponding tagged version is at revison 5789.
TRAC changeset: 7aa77565b437475c5ecd0efc1e174ca3976110c7
Action: In Progress
Tag ESM1.5 version of CABLE
4. Discuss next steps towards a minimum objective test suite and indicate timeline (?).
a. Institute Martin’s multi-config run script (I agree that “how” is open), and set up a corresponding PALS experiment for testing its results
b. Add ACCESS CM2 compile and (very short, say 1 week) execute to multi-config script, since this version of CABLE (in theory) runs with CM2
c. Add (short) global offline runs to the multi-config run script, and set up PALS experiment (ILAMB or otherwise)
Action: In Progress
- Martin to email community for ideas of key flags that need testing (e.g. MPI test priority)
Action: In Progress
- Martin & Gab to work on benchmarking example by next meeting.
5. Discuss the need to fix the CABLE offline output time variable. I think there are also issues with the time variable in general.
Action: In Progress
- This remains an issue with no one volunteering to fix it.
6. Discuss steps to get CNP working.
Action: In Progress
- Martin to set up test case, Ying-Ping to visit Sydney for two days in April to see if this can be progressed.
7. Discuss CABLE coding strategy in general. For e.g. There was an issue with Ying-Ping’s LUC and writing restart files that boiled down to a tension with “mp” format of CABLE vectors. In JULES this would not be
an issue as they use much simpler format.
Action: In Progress
- YP had a follow up discussion with Gab on the best way to fix this, they are working on this together to make the restart reading more flexible.
8. Any other business?
-
Discussion of land use change separation issue, no resolution reached.
-
Coding standards issue was raised. It was decided that moving forwards all new science tickets need to meet the Met Office protocol. Martin added a link to this on the Wiki
-
Issue around CABLE's open source licence and the potential conflict with the Met Office crown licensing arrangement was raised.
-
lakes and soil colour albedo parameterisation (#207)
-
temperature vs potential temperature (#206)
-
Some met variables 4D in site runs: Presumably this is required and this ticket should be rejected? (#200)
-
Tiled vegetation cannot be passed through met file (#199)
-
partitioning evaporation into evaporation and sublimation (#196)
-
clean up roughness variables (#153)
-
Allocation wrong in CMIP6-MOSRS when using MODIS phenology (#204)
-
Restart reproducibility (#198)
-
bare soil roughness and sparse vegetation (#205)
-
soil emissivity =1 incorrect (#203)