Skip to content

May2016

Jump to bottom
Gab Abramowitz edited this page May 30, 2016 · 8 revisions

CABLE committee meeting, 27th May 2016, 3.00pm

  1. Trunk updates:

a) Updates this month - none

Note that recent tests of current trunk indicate it is not correctly simulating carbon fluxes. Problem seems to have occurred with #2 implementation. Under investigation but reiterates need for basic tests across a range of applications when updating trunk.

b) Updates in progress

Little progress to report.

i) #62 (radiation/albedo bug fix, ASAP), #92 (if BP confirms), #66 (initialisation bug), #89 - Jenkins tests?

ii) SLI and POP and other VH/LN enhancements, #18, #61, #73, #93, #90, #94. RL reminded BP.

iii) #95 (and #44)

iv) #70 - BP/JS checked

Ready to go.

c) status of CABLE-AUX files e.g. cable.nml

Need to ensure all files on CABLE-AUX are up to date, e.g. all switches listed in .nml files.

  1. Potential CABLE community activities

a) Australian Academy of Science review (due 5th June)

Agreed that it is worth making a submission as a committee. RL to put some information into required form and then circulate for extra input.

b) MIPs associated with CMIP6, LS3MIP, LUMIP, ESM_SnowMIP - coordinate contributions?

Agreed that this would need a video conference or part of workshop to progress. Not urgent. Key need is to have online and offline versions consistent, so work cannot begin until online model agreed. Can start to get infrastructure in place. COE could likely contribute to LS3MIP given their past experience with GLACE type experiments. RL to send tentative yes to ESM_snowMIP.

  1. CABLE-JULES coupling

a) Progress with coupling into UM8.5 for ACCESS-CM2.

Little progress since last meeting.

b) CABLE in JULES4.2/UM10.1

CABLE prognostics enabled in I/O, CABLE calling points to be finalised.

  1. Benchmarking/test suite

a) CABLE technical test suite – Jenkins work (NH), ACCESS tests

Not much to report since last meeting. Agreed that undefined/uninitialized variable tests would be useful. Noted current test in codes for no change in result relative to previous version. Needs to ensure carbon fluxes are being checked as well. For Jenkins need to define a pass/fail threshold.

b) PALS

GA reported on discussions at ILAMB meeting around where PALS fits with ILAMB and LVT. After discussions with ILAMB Python developer and Sujay (LVT developer ) it seems both ILAMB and LVT could work as analysis engines in a PALS workflow environment with a relatively simple wrapper. PALS distributed architecture could then allow analysis to happen in the same location as data if a local worker node were installed (and push results to the web server where results/reporting display occurs). Would allow a worker node at each institution e.g. NCI. Generally felt to be a good idea, need to establish whether this is achievable.

  1. Meetings

a) UM users workshop - land surface session, June 14th

RL seeking input for CABLE status and plans talk. Sent email to list. Little response so far. May need to chase people.

b) AMOS conference 2017, special sessions

No indication yet on success of special session proposals.

  1. Next meeting date: Friday June 24th or July 22nd, 2pm

Still seeking feedback on preference, but likely July 22nd unless urgent tasks emerge.

  1. Any other business

Clone this wiki locally