ACCESS Oversight Committee, 19 August 2019


The issue of formal ACCESS Model Version Releases was discussed.

  • CM2 – CMIP6 configuration frozen. Martin Dix checking portability for non-p66 users on raijin with the intent to hand over to Scott Wales. Will be made available via Rose/Cylc suites for key experiments [AMIP, historical, future scenarios, DEK experiments]. Documentation will be via WIKI. Restart files will be made available. Key contact is Martin Dix.
  • ESM1.5 – code handed to CLEX, but new code fixes being tested and will be updated. AMIP configuration to be developed for carbon cycle version only. Restart files made available. Key person is Tilo Ziehn (unavailable until mid-October). Time line 1-2 months.
  • ACCESS high resolution regional model – none available for general use at present but BoM examining the possibility of providing a release of a high-resolution version of ACCESS.
  • Input on training needed for ACCESS requested. Please forward requests to institutional leads

CMIP6 update

  • 4 x CO2 and 1% runs completed. Scenario runs SSP585 and SSP445 done. Other runs on-going. CSIRO notes net TOA heat fluxes pretty good cf other CMIP6 models. Trends in global surface temperatures in PI control low in ACCESS ESM1.5, slightly higher in ACCESS CM2 [but CM2 run not yet complete, 350/500 years). Scenario runs look broadly in the middle of the pack.
  • Another land use bug in emissions-driven run in ESM runs has been found, requiring re-run.
  • CMIP6 data being downloaded based on demand and agreed prioritisation. A specific request for daily rainfall ~10 ensemble members from CMIP6 is being resolved.
  • Conversations between the NESP hub and NCI around some high impact visualisations for ACCESS for CMIP6 are on-going. If this interests you, please talk to your institutional lead.

Other general matters

HPC issues

  • New HPC system (Gadi) and subsidiary systems being transitioned to the new NCI machine. There are some issues in terms of transition of CM2 and ESM1.5, and making sure that the versions of CM2 and ESM1.5 brought across to the new HPC system are fully traceable.
  • Issues relating to /short on the new machine are noted and are being discussed with NCI.


Back to AOC meeting notes

Back to main ACCESS wiki page

Last modified 10 months ago Last modified on Mar 31, 2020 3:08:44 PM