Changes between Version 1 and Version 2 of access/AocOct2020

Dec 17, 2020 12:44:51 PM (3 years ago)
Rachle Law



  • access/AocOct2020

    v1 v2  
    1 = ACCESS Oversight Committee - Notes from the 24th September 2020
     1= ACCESS Oversight Committee - Notes from the 22nd October 2020
    3 == 1. Round-table - update on activities and issues relating to ACCESS
     3== 1. ACCESS-NRI
     5$7.6m has been allocated over 3 years starting on 1 July. Department has contacted ANU DVCR (Keith Nugent) as ANU is likely host. Early steps will likely be to appoint a director and a governance model made up of partners. Proposed A. Hogg contacts K. Nugent to make him aware of AOC if any input from this group is useful. This group could be replaced by NRI governance. It would be good if communities of interest organise themselves to reach common view of infrastructure needs.
     7== 2. Round-table - update on activities and issues relating to ACCESS
    59=== BoM
    7  * The hindcasts of the ACCESS-S2 seasonal prediction system are ongoing and approaching 50% complete.  The hindcasts are for the period 1981 to present, with 3 ensemble members to lead time of up to 9 months starting on each of 11 start days per month, with additional 6 ensemble members to lead time 6 weeks starting on each of 6 start days per month (with focus on the multi-week prediction).  The ACCESS-S2 real-time trial is to commence in the next month, with the aim of becoming operational around March 2021.
    8  * The Bureau has started runs of the UM-based downscaler at 2.2 km resolution over eastern Australia.  Will do two 20-year periods 1986-2015 and 2040-2059 nested in ACCESS-1.0.  This is a pilot project to ascertain efficacy.
    9  * There is a JSHESS special issue coming out on collaborative work between the Bureau (Huqiang Zhang lead) and CMA and Met Office researchers on 'atmospheric river' behaviour over the  Australian and east Asian region and its simulation in the UM.
    10  * The Bureau Annual R&D Workshop "A Land of Storms, Floods and Bushfires:  Seamless and Integrated Forecasting" will be held online 23-26 November 2020.
     11 * Charmaine replacing Tony on AOC. Tony thanked for his contribution.
    1213=== Universities (including STRESS2020 high res simulation)
    14  * CLEX workshop - 16-19th November
    15  * 400m runs to be announced and released early October
     15 * Working through research plans now that ACCESS-NRI decision known.
     16 * Mid-term review went well.
     17 * ACCESS CMIP6 versions training run on 20th October was helpful
     18 * Dataset from 400m simulation being prepared for release. Workshop likely.
    1720=== CSIRO (including ALCG simulations)
    19  * Working though ALCG simulations primarily on DAMIP runs
    20  * ESM runs now 30 members for historical and SSP-245 scenario
    21  * Issues with storage raised
     22 * The ACCESS lightning-produced NOx work is out in ACPD now: and the new lightning scheme has been committed to the UM trunk: see ticket
     23 * As part of the Australian Antarctic Program Partnership, work is being undertaken on improving the Southern Ocean radiation bias in the UM by improving the representation of marine aerosol and cloud processes. We are collaborating closely with UK Met Office, NIWA, and others on this activity.
     24 * AMOS 2021 conference workshop on ACCESS-CMIP6 models is expected to go ahead assuming sufficient interest
    2427=== NCI
    25  * CMIP publication of data proceeding
    26  * ACCESS code – performance evaluation on-going
     28 * CMIP publishing taking significant time. NCI contribution is much appreciated by CSIRO.
     29 * Visualisation progressing. Discussion on whether this can be more coordinated across ACCESS (e.g. video that shows range of applications). Noted that included in NRI.
     30 * ERA5 – announcement of user access not far off – early November. Takes time to get data from ECMWF.
     31 * HPC roadmapping / NCRIS – NCI asked to do some more work/engagement – so good to be clear about what we need for our models.
     32 * Pawsey announced next machine, AMD CPUs and GPUs. Late 2021-2022 delivery. Not much storage. Couldn’t run ACCESS on GPUs, not sure about balance of CPUs.
    2835=== COSIMA
    29  * OMIP runs with ACCESS OM2 @ 1 degree with WOMBAT.
    30  * Several additional experiments
    31  * Possible 0.25 degree OMIP (physics only)
    32  * 10th degree model also running for comparison.
    33  * Storage issues at NCI problematic
    34  * Monash funding for storage via LIEF for ERA-5 and model output arrived at NCI
    35  * Notes BoM workshop on the future of ocean modelling not leading to convergence.
     36 * OMIP runs published soon – not quite same version as ESM1.5/CM2 – so new institution-id/model-id for ESGF. Also publishing IAV forced 1/10 model. No news yet on COSIMA2.
    3738=== NESP
    38  * Planning ACCESS CM2 and ESM1.5 webinar on (provisionally) 20th Oct. Update on CMIP6 data sets too.
    39  * Delay on NESP2.0 announcement – delayed until “towards the end of 2020”. Start date uncertain.
     39 * A “Getting Started with ACCESS-CM2 and ACCESS-ESM1.5” (CMIP6 configurations) training session was held (20 Oct) for potential new model users. There were 64 attendees from across CSIRO, BoM and universities. A recording of the presentation, along with documentation and the meeting slides are available via the NESP Earth Systems and Climate Change Hub website
     40 * CMIP6 ESGF downloads:
     41  * ACCESS-ESM1.5       +2M downloads, ranked 15/135 CMIP6 models
     42  * ACCESS-CM2   +1M downloads, ranked 25/135 CMIP6 models
     43 * NESP ESCC science symposium 8-9 Dec around research achievements from the Hub. Partly face-face (1 site per State). [did not go ahead]
     44 * Consultation draft for Next Gen projections – targeted distribution for feedback. Collated information (e.g. across CLEX) useful.
     45 * David K – last meeting. Thanked.
    4149=== UM Partnership activities
    42  * UM Virtual workshops – 9-13th Nov (7-10 pm)
    43  * GC4 assessments available for examination – see
    45 == 2.   ACCESS related training and webinars (plans, requests)
    47  * Addressed above
    48  * Half day meeting associated with the AMOS meeting
     50 * Each partner has nominated at least one Next Generation Modelling System (NGMS) champion to facilitate planning of uptake/transition to NGMS from UM. Local champions are Justin Freeman and Joerg Henrichs (BoM) and Peter Dobrohotoff (CSIRO).
     51 * UM Users workshop is Nov 9-12 (evenings for Australia)
     52 * Given surplus UM Partnership fees (no travel), the Partnership will fund secondments for LFRic transition support and Evaluation Tool/software portability. Partners have been asked to propose suitable staff who could make full or part time contributions over 1-2 years.
     53 * GC4 Assessment workshop will be taking place 2-3 Dec 2020,
     54 * Interaction of UM partnership with ACCESS-NRI will be important as ACCESS-NRI is developed.
    51 == 3.   ACCESS development planning (including but not restricted to NESP2 and/or ACCESS-NRI)
     57== Other business
    53  * Wait to see outcome of the ACCESS NRI
    55 == 4.   Any other business
    57  * WCRP rethinking plans around “Lighthouse activities” in the future. Interwoven with new technology. Early planning around writing science plans.
    59  * Next meeting 22nd October
     59 * Next meeting 10th December