Changes between Initial Version and Version 1 of access/AocApr2020

May 5, 2020 3:49:38 PM (6 months ago)
Rachle Law



  • access/AocApr2020

    v1 v1  
     4= ACCESS Oversight Committee - Notes from the 30th April 2020
     6== 1. Round-table - Update on activities and issues relating to ACCESS
     7=== CSIRO
     8 * CMIP6 progress: CM2 now completed 3 member ensembles for 4 Tier 1 SSPs (ESM1.5 completed these previously). CM2 RFMIP (AMIP style) runs completed. Backlog of processed data waiting and will then be published through the ESGF.
     9 * Multiple accepted papers: Ziehn et al. (ESM1.5 model description), Bodman et al. (AMIP evaluation), SIMIP (sea-ice) multi-model papers.
     10 * Gadi: Significant speed-up of ESM1.5 with extra cores for ocean (2h15m/yr to 1h25m/yr). ESM1.5 payu version on gadi not yet working (bit reproducibility issues)
     11 * Comms: ACCESS page on updated and slowly adding material. Staff liaising with NCI on ACCESS CMIP6 visualisations
     12 * Discussions with aerosol team on using ACCESS for GBR cloud brightening project
     13 * COSIMA2 ARC Linkage bid submitted - led by ANU
     14 * Decadal team working on data assimilation and COSIMA 0.25 degree ocean. Also working with BoM on sea-ice DA.
     16=== BoM
     17 * Trials of city-scale models with 1.5 km resolution and 12 member ensemble at 2.2 km resolution. Operational in a month or two.
     18 * Running a summer hindcast at this resolution (Dec to end January) to cover period of bushfires and hail storms. If useful may move data to NCI for wider use.
     19 * Plans for polar modelling, Antarctic region to support new airport. Considering boundary layer and physics. Focus is on 24 hour forecasts.
     20 * Accessdev issues have been hindering ability to use gadi.
     22=== CLEX
     23 * Tipping points workshop run with paleo focus. Potential greater use of ACCESS-ESM1.5 but would need further development
     24 * Antarctic initiatives (Monash led and UTas led). If significant ACCESS use then may want to join AOC. Andy P to contact Matt England. Christian to talk to Monash lead.
     26=== NCI
     27 * Finalising Climate DeVL report – will be circulated
     28 * Rolling out data analysis tools for CMIP6 (Jupyter notebooks) including ‘How to use’
     29 * Preparing for updated cloud-based infrastructure – aiming for later this year
     30 * ERA5 ARC LIEF grant led by Monash successful. Setting up governance group, covering use for initialisation and data analysis. Including tools.
     31 * Working with BoM for better management of datasets
     32 * Performance evaluation. Investigating power/vectorisation options for codes on Gadi processors  – will provide advice on compiler options etc.
     34=== NESP ESCC
     35 * has compiled comparison of performance from Raijin to Gadi for climate cases.
     37=== Successful ALCG application for ACCESS climate modelling
     38 * 37 MSU for next 3 quarters for ACCESS-ESM1.5, CM2 and chemistry simulations. ESM1.5 and CM2 focussed on additional CMIP6 runs: Tier 2 runs, long-term extensions, extra ensemble members and DAMIP. All datasets will be available for community use. Contact Rachel Law if interested in other CMIP6 runs.
     39 * Decadal team grant for large ensemble, open for community use.
     41== 2.   Release of ACCESS-CM2 CMIP6 version on Gadi
     43 * Would be good to find a potential test user (with little experience)
     44 * Work continues to ensure robustness and include netcdf conversion in suite. Martin encouraged to work with Scott, Ben, Holger to help with timely progression.
     46== 3.   On-line science meetings – assessment of interest – Rachel Law
     47 * Proposed an online meeting for CMIP6 multi-model and ACCESS work. Noted UK CMIP6 meeting worked well. Noted a few CMIP6 papers accepted that could provide talks. Also short talks on work in progress to avoid duplication of effort.
     48 * COSIMA doing webinar workshop shortly – so could see how that goes.
     49 * Suggestion to record sessions if timing not convenient for all (also some staff finding too many meetings).
     51== 4.   Grand Challenge Simulations – Update on STRESS2020 NCI initiative – Christian
     52 * Completed 400m resolution simulation for 2.5 days. Interest in doing more simulations e.g. 2.2 km for 1 year across Australian domain. Christian welcomes feedback / ideas.
     53 * Anyone know of a prize for technical support?
     55== 5.   NESP2
     56 * Jaci Brown is giving talks to CSIRO, BoM and Universities on NESP plans
     57 * Dependency of future ACCESS work on ACCESS-NRI
     58 * Dependency on NCI
     61== 6.   Any other business
     62 * AOC notes will now be put on the ACCESS wiki - can be distributed across institutions.
     63 * Evaluation tools: Noted a tension between providing standardised tools (with support/investment) and individual researchers developing their own methods/tools. University users seem happy with their current tools. Agreed to wait and see how BoM discussion develops.
     64 * UM users workshop in June will be 4x1 hour sessions, one each day Monday to Thursday (7PM AEST). See recent UM newsletter for more information and registration link.
     66Next meeting: 28th May. Choice of video conference system left to chair (Peter Steinle). Ben Evans (rapporteur).
     73[wiki:access/AocMeetingNotes Back to AOC meeting notes]
     75[wiki:access Back to main ACCESS wiki page]