ESM Working group: Meeting notes 2026

Earth System Working Group
, ,
1

Notes from each meeting of the ESM Working Group are added as a reply to this topic.

Make each reply a wiki post, that way other attendees can edit the notes. This can greatly reduce the burden of note-taking allowing that person to participate fully in discussions without worrying too much about missing important details.

Archived meeting notes:

2023
2024/2025

3

Date: 3/2/2026
Chair: @tiloz
Participants: 16

1. Admin and resource usage

  • Welcome back: Welcome back to the ESM WG meeting series. As discussed towards the end of last year, meetings have now fully swapped to the Tuesday 10:30AM schedule, and will alternate each fortnight with the CM3 evaluation and development meetings. Make sure to follow this topic , as announcements for both types of meetings will be made there.

  • Science Presentations: We don’t have any science presentations lined up for this year, and so if you have any work that you would like to share during the ESM WG meetings, please get in contact the ESM WG co-chairs or @spencerwong. Presentations from recent conferences and workshops are welcome, as are informal updates on in progress work.

  • LG87 Resource Usage:

    • Compute allocations for each working group have been doubled to 1.5MSU this quarter. 1MSU of the ESM WG’s allocation has been assigned to proposed experiments, leaving a large amount available for additional proposals. With the increase this quarter, there may be large amounts of compute available for rebalancing from other working groups at the end of the quarter, and so if you have any experiments to run, it’s a good time to put up a proposal and get them running.
    • If you have any experiments you are interested in running with ESM WG resources, follow the proposal guidelines here to apply to use WG compute.
    • Storage usage at 65 TB of 100 TB allocation.

  • Next CM3 dev-eval meeting cancelled: The CM3 dev-eval meeting scheduled for 17 February has been cancelled as it clashes with AMOS. There will be an ESM WG meeting two weeks later on 3 March followed by a CM3 dev-eval meeting on 17 March.

  • Shared calendar invite for meetings: The shared calendar for ESM WG/CM3 Dev-eval meetings has now been updated. If you were already subscribed, the meeting invitations should now have the correct times (contact @spencerwong if there are any issues).

    If you aren’t already subscribed, please follow the instructions on the how to subscribe to a calendar feed page using the calendar URL https://outlook.office365.com/owa/calendar/463c5a6d0bba4c838dce9c57933398fb@anu.edu.au/3086e7a51b0d48ef85bee981a597274413647516456690471131/calendar.ics to add the shared calendar to your preferred application.

    The above instructions will be added in a post on the announcement page.

3. ESM1.6 updates

  • @tiloz and @RachelLaw presented updates on the ESM1.6 simulations completed over the holiday period. Key updates include the resolution of the ocean temperature drift with the ocean albedo decrease, updates to all CMI7 forcings other than ozone, and performance improvements compared to ESM1.5, with a ~40-45% increase in throughput for the same compute cost.
  • Updated configurations tested over the holiday period include: concentration driven picontrol, emissions driven picontrol, fully coupled, bgc-only, and radiation only 1%CO2 simulations, 4xCO2, flat10, and historical simulations.
  • An issue was identified in the land BGC: BGC only 1%CO2 simulations did not show the expected magnitude of increase in land carbon uptake. This has been attributed to a bug in the phosphorous leaching parameter, which depleted the phosphorus pools available to plants. This bug has now been fixed, and new simulations with increased initial pools are being run. Carbon-nitrogen only simulations are also being tested, which display the expected levels of carbon uptake.
  • Historical simulations produce a similar climate to ESM1.5, however the land carbon cycle is quite different. Several tests are being run with different settings for phosphorous limitation, wood harvests, and initial pool sizes, to help with tuning the configuration.
  • Outputs from these test simulations are available for analysis in the following locations:
    • December spinup (CNP): /g/data/p73/archive/CMIP7/ACCESS-ESM1-6/spinup/Dec25-PI-control
    • December spinup emissions driver (CNP): /scratch/p66/txz599/access-esm/archive/esm-pi-01-expt-4a4edd44
    • December spinup (CN): /g/data/p73/archive/CMIP7/ACCESS-ESM1-6/spinup/Dec25-PI-CN
    • Spinup test with phosphorous bug fix (CNP): /scratch/p66/txz599/access-esm/archive/test_pi_plss_02-expt-a9acd03a
    • 1%CO2 (CNP + phosphorous bug fix): /scratch/p66/txz599/access-esm/archive/test_1pct_04-expt-2cb54419
    • 1%CO2 (CN): /scratch/p66/txz599/access-esm/archive/testCN_1pctco2-expt-0bf72499
    • 4xCO2 (CNP with phosphorous bug): /scratch/p66/txz599/access-esm/archive/4xCO2-01-expt-822dff3c
    • Historical (CNP with phosphorous bug): /scratch/p66/rml599/access-esm/archive/hist-restart170-hist-restart170-78be432c
    • Historical (CN): /scratch/p66/rml599/access-esm/archive/test-hist-CN-test-hist-CN-3e80d641
    • Historical (CNP + phosphorous bugfix): /scratch/p66/rml599/access-esm/archive/test-hist-fixedP-test-hist-fixedP-4af5fa4d

Please feel free to correct any mistakes in these notes directly, or message @spencerwong with corrections.

Additional information

4

Date: 17/3/2026
Chair: @ctychung
Participants: 16

1. Admin and resource usage

  • Science Presentations: Spaces are regularly available for science presentations at ESM WG meetings. If you have any work that you would like to share with the community, please get in contact the ESM WG co-chairs @dkhutch, @tiloz, @ShayneM, @ctychung or @spencerwong. Presentations from recent conferences and workshops are welcome, as are informal updates on in progress work.

  • LG87 Resource Usage:

    • 1.09 MSU out of the 1.5MSU compute allocation has been used. The remaining allocation is expected to be used by proposed experiments.
    • If you have any experiments you are interested in running with ESM WG resources next quarter, follow the proposal guidelines here to apply to use WG compute.
    • Storage usage at 65 TB of 100 TB allocation.

2. CMIP 2026 community workshop

  • @ctychung presented highlights from the 2026 CMIP community workshop, including an outline of the CMIP7 future scenarios, community MIPs, the launch of the Rapid Evaluation Framework, and changes to the data conventions compared to CMIP6. A scientific highlight was the analysis of high resolution models, and the effects of resolution in reproducing observed trends. Discussions during the workshop raised questions around “operational” CMIP activities as well as the role of AI/ML.

3. Technical presentation: ESM1.6 restart creation

  • @lachlanswhyborn presented on new tools for creating ESM1.6 restarts for new land cover data.
  • In ESM1.6, the CABLE land model discretises each grid cell into tiles, which represent different surface types. Each tile has its own physical state, carrying data for variables like temperature, soil moisture, and nutrient pools. When creating a restart file which uses a new distribution of surface types, new tiles don’t automatically start off with valid data. They need to be initialised with valid and meaningful values, otherwise the model may crash or start far from equilibrium.
  • The restart creation tool provides new tiles with meaningful data for variables associate with the land such as temperature and moisture, and variables associated with the vegetation such as nutrient pools.
  • For land associated variables, surface fraction weighted means are calculated and used within each grid cell. For vegetation associated variables, a staged tile search is used to ensure that only data from relevant vegetation types are used to fill in the missing data.
  • This tool was used to set up ESM1.6 restarts for CMIP7 which used the new land cover datasets.
  • The tool is available at GitHub - ACCESS-NRI/esm1.6-scripts: Scripts and small programs related to ACCESS-ESM1.6 model and configs · GitHub, including documentation of its configuration options.

4. CMIP7 and CSIRO updates

  • @RachelLaw shared figures from the latest ESM1.6 simulations, which are expected to be CMIP7 production runs. Control, 4xCO2, 1pctCO2, and flat10 simulations are all running, and work is underway to tune the emissions driven historical simulation.
  • There hasn’t been a commitment to run the DAMIP simulations with ESM1.6 for the fast track. CSIRO likely does not have capacity to, however if you would be interested in running these simulations and have capacity to do so, please reach out to @RachelLaw.
  • CSIRO is currently facing job cuts, with a large number of cuts going to the Environment Research Unit. This will include job losses among staff working on ACCESS and ESM WG community members. There is currently a 3 week consultation, members raised whether there were avenues for the community to provide feedback on the proposal.

5. Enso focused group

  • @Dietmar_Dommenget is setting up an ENSO focused group for discussing all things ENSO related, including ENSO performance in the ACCESS models. If you are interested in being part of this group, click the bell to follow this topic on the ACCESS Hive ENSO - ACCESS Hive Community Forum, and reach out to anyone you know who may also be interested.

Please feel free to correct any mistakes in these notes directly, or message @spencerwong with corrections.

Additional information

5

Date: 14/4/2026
Chair: @dkhutch
Participants: 13

1. Admin and resource usage

  • Science Presentations: Spaces are regularly available for science presentations at ESM WG meetings. If you have any work that you would like to share with the community, please get in contact the ESM WG co-chairs @dkhutch, @tiloz, @ShayneM, @ctychung or @spencerwong. Presentations from recent conferences and workshops are welcome, as are informal updates on in progress work.

  • LG87 Resource Usage:

    • 435 kSU out of the 2MSU compute allocation has been used. The working group compute allocations have been increased this quarter, and there is an additional pool available which will be distributed to the working groups on a first-come-first-serve basis. If you have experiments you are interested in running, it is a good idea to add a proposal early in the quarter as this helps with resource allocation.
    • If you have any experiments you are interested in running with ESM WG resources next quarter, follow the proposal guidelines here to apply to use WG compute.
    • /g/data/lg87 recently came close to its 100TB limit. Recent cleanup has reduced the usage to 84TB, however, further cleanup of existing data including removing unused data and compressing outputs would be much appreciated. Freeing up space will allow for members to run new experiments.

  • Are you interested in running experiments for the CMIP7 fast track:

  • There are fast track experiments that CSIRO isn’t committed to running, such as the abrupt-0p5CO2 and aprupt-2CO2. Please see this post for a table of fast track experiments and the institutions assigned to run them. There is an opportunity for working group members to run unassigned experiments using working group resources. If you would be interested in running any such experiment, please get in contact with @RachelLaw.

2. Science presentation: Flux Corrections in Climate Models

@Dietmar_Dommenget presented his work on applying flux corrections in coupled climate modelling.

Method:

  • Flux corrections are commonly used to reduce biases a model’s mean state. For example to reduce biases in the SST mean state, and additional net heat flux can be provided to the ocean model. The added fluxes are time and location dependent, but don’t depend on the model state.
  • Flux corrections have previously been used to correct 3D atmospheric temperature, moisture, winds, land surface temperatures, ocean salinity and more.

History/examples:

  • Flux corrections were used in first generation GCMs to reduce biases in slab ocean SSTs.
  • Later coupled models with full ocean GCMs removed flux corrections due to them being viewed as artificial
  • Flux corrections have been used in several non CMIP areas such as seasonal forecasting, improving tropical cyclone forecasting, investigating SST warming patterns.
  • Examples from Dietmar’s research on flux corrections include investigation into the impact on mean state SST biases on global warming responses, and comparison of tuning and flux corrections as approaches for improving models.

Why use flux corrections:

  • Many feedbacks and processes are mean state dependent, meaning that flux corrections can improve model performance.
  • Model tuning can also be seen as artificial, may not perform as well, and is costly to perform

Example: ENSO performance in the ICON model

  • The ICON-XPP coupled model has a strong cold tongue bias and large biases in the Planton ENSO metrics. Dietmar has been investigating whether these biases can be improved via heat and wind stress flux corrections.
  • Flux corrections significantly improve the variability and processes metrics, including for example the equatorial mean SST, zonal mean precipitation, sst skewness, taux-SST and SSH-taux regressions.

3. UK Met Office user license agreement

  • Members from ACCESS-NRI partner universities ANU, UNSW, UTas (including AAD), Monash Uni and UniMelb who run ACCESS-AM, ACCESS-CM, ACCESS-ESM, ACCESS-rAM, or ACCESS-rCM will need to read and accept the new Terms of Use for the Agreement by 8 May 2026 in order to continue to use and access UKMO software and licensed material.
  • Please find instructions on how to do this and frequently asked questions here: Accessing UKMO licensed models.

4. ACCESS Community Workshop on Land and Coupled Modelling

  • The ACCESS Community Workshop on Land and Coupled Modelling will bring together the ESM and Land community working groups and will be held 31 August - 2 September in Melbourne. Abstract submission forms for the workshop will be out shortly.

Please feel free to correct any mistakes in these notes directly, or message @spencerwong with corrections.

Additional information

6

Date: 9/6/2026
Chair: @tiloz
Participants: 13

1. Admin and resource usage

  • Science Presentations: The last couple of meetings have been cancelled due to there being no science presentations. Several spaces are available for presentations at future ESM WG meetings. If you have any work that you would like to share with the community, please get in contact the ESM WG co-chairs @dkhutch, @tiloz, @ShayneM, @ctychung or @spencerwong. Presentations from recent conferences and workshops are welcome, as are informal updates on in progress work.

  • LG87 Resource Usage:

    • 1.88 MSU out of the 2MSU compute allocation has been used. The working group compute allocations have been increased this quarter, and there is an additional pool available which will be distributed to the working groups on a first-come-first-serve basis. If you have experiments you are interested in running, it is a good idea to add a proposal early in the quarter as this helps with resource allocation.
    • If you have any experiments you are interested in running with ESM WG resources next quarter, follow the proposal guidelines here to apply to use WG compute.
    • /g/data/lg87 has reduced from 94 TB to 63 TB out of the 100 TB allocation. Many thanks to everyone who cleaned up data. If you have any data in /g/data/lg87 it would still be much appreciated if you can clean up anything that is no longer required.
    • If you are interested in using working group resources to run CMIP7 experiments, please get in contact with @RachelLaw. Please see this topic for a list of the available experiments.

2. ESM1.6 and CMIP7

  • ACCESS-ESM1.6 production runs for CMIP7 are underway, with focus now shifting from the control and idealised simulations to the historical experiments. An overview of the status of the different experiments is available here. You are welcome to look at the output and get involved with the ongoing discussions and analysis.
  • A CMIP7 meeting on coordinating the analysis and evaluation effort was recently held. A future catch up could be hosted during the WG meetings.

Status of ACCESS-ESM1.6 experiments - CMIP7 / CMIP7 MIPs and other experiments - ACCESS Hive Community Forum

3. ACCESS Community Workshop on Land and Coupled Modelling

  • Abstract submissions for the ACCESS Community Workshop on Land and Coupled Modelling close at midnight AEST June 15.** If you would like to submit an abstract for a presentation or poster, please submit your abstract here.

4. Science presentation: Missing Antarctic meltwater: Implications for Southern Hemisphere climate projections

@wghuneke presented on the impact of Antarctic meltwater on the Southern hemisphere in coupled climate models.

Background:

  • The Antarctic ice sheet and its interactions with the ocean are not well represented in current climate models. For example, basal meltwater is typically prescribed and lacks variability and feedbacks.
  • Recent climate model experiments show that there is large uncertainty around the effects of Antarctic meltwater on the climate.
  • The SOFIA freshwater experiments comprise of 6 idealised meltwater experiments, with the aim of allowing for multimodel analysis. This presentation discusses results from the SOFIA antwater experiments, where 0.1 Sv of freshwater is added around the Antarctic coastline.

Results:

  • Ocean: The added freshwater consistently causes surface cooling, subsurface warming, and a reduction in Antarctic bottom water formation which varies across the model. The magnitude and pattern of surface cooling also varies across the models, however doesn’t correlate with the magnitude of the changes in deep convection
  • Sea ice: Sea ice area increases in all models, with magnitude and spatial variations depending on the models’ mean states.
  • Atmosphere: Cooling in the troposphere is strongest around Antarctica, however spreads to the northern hemisphere. Strengthened westerlies are seen in the southern hemisphere.
  • Tropics: The meltwater produces a spatial pattern of cooling in the tropics and leads to a northward shift of the ITCZ.

Impact on southern hemisphere land masses

  • Compared to the SSP2-4.5 experiments, the antwater experiments largely produce opposite signed responses in temperature and precipitation. The additional meltwater is expected to damp temperature increases over southern hemisphere land masses. nLarge spreads occur within each model, and across the models, emphasising the importance of using multiple models and as large ensembles as possible.
  • Precipitation responses are generally not significant for the current number of ensemble members, and the models do not agree on the sign of the changes, with the only robust response being drying over Antarctica.
  • The magnitude of cooling and drying over Antarctica strongly correlate with the change in air temperature over the southern ocean, highlighting potential feedbacks onto melting.

Please feel free to correct any mistakes in these notes directly, or message @spencerwong with corrections.

Additional information