ESM working group: Meeting notes 2023 (Archived)

Earth System Working Group
, ,
1

This is an archive of the ESM Working Group Meeting Notes from each meeting of the ESM Working Group from 2023.

2

Date: 27/04/2023

Participants: 20

Working group goals:

  • Platform for Australian researchers to have input in versions and configs used by community for research
  • Currently based on configurations of ACCESS-CM2 and ACCESS-ESM1.5
  • Also based on applications and research done with models. From paleo to future projections.
  • Build a community on research.
  • Co-develop future versions.
  • Set up new collaborations and strengthen existing collaborations across projects and organisations.
  • Allocate ACCESS-NRI compute (1.25 MSU per quarter) and some storage for shared work/goals

Ran through introductions of participants, who they were and their research interests.

Q: What is role in ESM Working group in coupled model develeopment? Working group can set priorities and work on some configurations and pass back to ACCESS-NRI.

Q: Who is responsible for coupled model development work in ACCESS-NRI if there is no coupled model group? Good question, and something that is being worked through. Will involve collaboration between modelling groups and be governed by priorities of model release.

Move to CMIP7, group can be important to prioritise which models to continue to support, and where development needs to go.

Keep ESM1.5 or need a faster new model? Dietmar flagged requirement for 100 model years/day. Also currently working on UM with N48 atmosphere. Paleo needs to model carbon cycle. If possible dynamic land surface. Don’t care so much about atmospheric resolution.

Future of marine carbon cycle?

CM2 and ESM1.5 both have their uses.Need multiple configurations. Not just one model. WG can be a forum to discuss what configurations are required.

@wghuneke working with @MartinDix on CM version with 0.1 deg ocean, but obviously slower, and still ironing out some errors with the 0.25 deg ocean CM version.

Map out intersection points of interest. e.g. land/sea masks and ancillaries. Has been challenging, pool resource and improve processes.

Land-sea masks key step for changing coupled model configurations. Knowledge is limited to a few people. Can ACCESS-NRI form a team to do this work. Create package to do this usable by community.

Re: meeting frequency: Monthly maybe too far apart. If one is cancelled the gap is too big.

Have a focus on each meeting, allows those who are keen to try and make it. Also good to keep it diverse to learn from others.

Need good support for calendars to allow meetings to be easily added.

Decided on fortnightly meetings, but first meeting is scheduled in a month for 2023-05-25T05:00:00Z.

This is a wiki post, if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me

3

Date: 25/05/2023

Participants: 16

Chair: @ShayneM

Recap:

Decided on fortnightly alternating 1pm/3pm until @tiloz returns. Include a science talk to stimulate discussion. Email Co-chairs or @Aidan if you want to talk, include a timeframe for when you would like to talk.

VC: Zoom is ok for BoM, so continue

Some issue with ESM Calendar (1pm meetings missing) @Aidan fixed.

ACCESS-NRI Workshop

@dkhutch and @ShayneM said they would be happy to help with Program Committee, and @LaurieM would step in if required.

Need to know if people who attend this meeting would attend an ESM WG satellite meeting of the Workshop, or attend one of the other WG meetings as there is significant overlap.

Once timing is settled could use a poll on the hive for working group: who will attend WG day?

What’s going on?

Who is running what? Need a place to put details of existing runs. Need knowledge of what has already been run as context for distributing resources.

Ariaan has freshwater perturbation runs for southern freshwater initiative. Used ACCESS-ESM1.5. PI-control perturbation. Planning similar runs ensemble with different spatial distribution. Previous ensemble was 5x100 years. BGC output is potentially interesting to others (Laurie). problem is resources for someone to analyse the data. Also depends on size of perturbation. Need to perturb physics enough: 0.1 Sv so might be detectable. 1 degree BGC isn’t huge in terms of extra resources.

@LaurieM: Also looked into making more use of 2 carbon tracers. Radiative forcing fixing co2 to 284, also interactive CO2 tracer, atmospheric tracer reacting to carbon cycle. Will do interactive CO2 for glacial experiments. Could include this tracer in freshwater perturbations.

@tiloz: currently doing LESFMIP runs. Extension of AMIP but focussing on large ensembles. Will be published on ESGF. Quite a number of runs, good to have feedback from community on which experiments to do first. Currently only looking at tier 1 runs, but could do others if strong interest. Currently doing 7 runs: historical period, volcanic forcing only, land use change only, ozone forcing. solar forcing only. Currently have 10 ensemble for each. Aiming for 40 gradually. Can prioritise which to do first. Total ozone forcing. Different to the DAMIP? Should be AMIP protocol. Runs under CSIRO. Will do this gradually when compute time is available. Happy to share location of data.

@dkhutch: wanting to use ACCESS-ESM1.5 to do miocene/eocene experiments. Dietmar and Abhik have similar goal. Having trouble changing properties such as land/sea mask, vegetation and boundary conditions. They asked for GFDL CM2.1 as interim step. Will make a user story on the forum how that was done, serve as a template for ACCESS-ESM1.5 process.

@paulleopardi: looking at ancillary data for ACCESS. Would like to have your use case.

@jbrown: David implementing PMIP or DMIP? MioMIP 15MYA. Needs extensive changes to all boundary conditions. Want to do for pliocene, but seemed too hard. Have to do something similar for glacial experiments. Dramatically perturb boundary conditions. Some experience in Dietmar’s group. Have simulations with inactive land surface (proscribed land surface). Might have some useful experience. Dietmar has been able to move some land masses in N48 with slab ocean. Need to extract that information.

@atteggiani working with @Dietmar_Dommenget and @abhik on ancillaries and land/sea mask. Move mask ok, but need sensible vegetation for land types. Get in contact with @atteggiani.

@MartinDix and @atteggiani will be talking next meeting about the progress with ancillary tools.

@LaurieM did change veg cover with last inter-glacial. Confident can do again.

@RachelLaw: mapping out model dev pathways we might take for CMIP7. Could discuss at another meeting. @Aidan adding this to next meeting agenda.

Do we need to prepare or plan for CMIP7 discussion? Thinking about who is working on what and possible development pathways.

Will become a Hive thread after some consultation. Important to understand bottlenecks and where to prioritise effort. Different pathways could take, which one dependent on smoothly technical dev proceeds, evaluation of performance, drop back to fallback position.

Shared Experiments

lg87 is project for ESM WG. 1.25MSU/quarter. Want good cutting edge experiments, enhance science, more people interested the better.

Have a month to do something with the resources.

Aidan made resource allocation guidelines and example proposal on Hive (posted links in chat).

Use it or lose it.

Science talk

Tilo Ziehn: Implications of net-zero emissions at different global warming levels. ACCESS-ESM1.5 millennium length simulations.

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me

2 Likes
4

Date: 08/06/2023

Participants: 15

Chair: @LaurieM

Admin

Change next meeting to 3.30pm. Science talk volunteers.

AMOS 2204

AMOS Session proposal deadline tomorrow. Some already part of proposals. @Harun_Rashid, @tiloz, Ian Patterson and @eunpalim organising coupled modelling session past few years. Happy to work with volunteers from ESM Working Group. Will co-organise. Any volunteers from WG, let Eu-Pa know, today or early tomorrow morning.

Working group resources

No-one has volunteered to use ESM WG compute resources with 3 weeks to go. Anyone need compute resources this quarter, or ideas for next quarter. Organise and develop proposals around shared experiments useful for community. @tiloz has runs which could use up remaining resources. @Aidan contact Tilo and see if he can us use it

@ShayneM will talk to Arnold about running more members in CM2 historical and SSP members.

@LaurieM Should put proposals on hive, transparent, might encourage others to suggest experiments. ACCESS-CM2-025 runs starting soon. CM2 with WOMBAT. Dave Bi is leading this.

Reply to this topic to discuss using the available resources:

ACCESS-NRI Workshop

@dkhutch attended program committee meeting. Structure of workshOP

  1. Mon: Training day
  2. Tue: Main workshop
  3. Wed: Main workshop
  4. Thu: Working group day
  5. Fri: Optional second working group day

Does ESM WG want a second day? Probably only a day required.

Presentations: want science and model development presentations. Leaning towards science in posters and oral presentations more likely to be model development/software development. Each WG will be given a small number of oral presentation slots. Want to make talks as relevant as possible to entire community. Also would like a fun activity to build team spirit. Like COSIMA who went sailing.

Push for model development/software is from workshop committee. All WG have meetings on the same day, there will be overlap in working groups. Difficult to solve.

Poll who is attending? Who attend ESM WG day? Who wants to present on WG day? Who is planning on attending a working group meeting, and which one?

@Aidan organise poll.

Could have science talks during WG meeting. Discuss large experiments as a group.

If have ideas reach out.

@wilma will attend Program Committee meeting on 9th June.

Ancillary file work

UM Ancillary file introduction from @MartinDix.

Slides: Ancils_ESM_WG.pptx (794.9 KB)

And a presentation from @atteggiani

@paulleopardi is documenting existing ancillary files at UK Met Office wiki (account required)

https://code.metoffice.gov.uk/trac/TI/wiki/SIDAtNCI

Perhaps have a dedicated session to walk through an example? Moving Australia is a an example we could do. No general tool available. Organise a special meeting in July?

@paulleopardi Is HadCM3 a useful example?

http://www.paleoclim.org/methods/

@dkhutch Yes Paul Valdes has done these kinds of changes extensively for the HadCM3L model. It might be a good idea to ask him for his scripts (if he is willing to share) for comparison. Pretty much all the Bristol paleo simulations are using scripts from Paul Valdes to generate their boundary conditions

@atteggiani worked through example of altering vegetation ancillary using his modify_ancil tool.

@martindix: Anything that doesn’t change the land-sea mask are fairly straightforward.

@dkhutch: Confusing number of orography components in an existing ancillary files. Looked like more than one land-sea mask variable. Need more information about which variables are used.

@Dietmar_Dommenget: Big question is how do we change the land-sea mask. No idea what the timeline is. @MartinDix: atmosphere could be done in a week, not sure about the ocean side.

@Katrin: know what we want to do with topography and vegetation, ocean bathymetry and land-sea mask. Would provide. Question is how to translate them. Not everything UM needs we don’t have, e.g. standard deviation of orography. @MartinDix: sub-grid scale orography and mapping paleo onto CABLE vegetation type. @LaurieM has done changes in vegetation without problems. Did it for the last interglacial. Can change the files that are there. Should be able to change topography by merging with existing file.

@MartinDix: Opening and closing a strait should be straightforward.

@sofarrell: Will need to make sure ocean climatology works with new bathymetry. May need long ocean spin ups.

@paulleopardi Perhaps changes can be done to the MASTER files and then the ancillaires re-genrated to make them consistent? There is a controlled vocabulary for the fields. See Met Office Codes Registry : um for the fields metadata.

Make a hands-on meeting to work through how to do land/sea mask changes in ACCESS models. @dkhutch Put something on the forum. Minimal perturbation to land-sea mask for ESM-1.5 as an example. See that procedure written down.

@MartinDix Will set up an example. Maybe close Bering Strait and open Panama Strait to show examples creating and destroying ocean cells.

Could do training on this on training day for Workshop, or ESM WG meeting.

Science talk

Ran out of time. Move Eun-Pa talk to next week.

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me

1 Like
5

Date: 22/06/2023

Participants: 22

Chair: @tiloz

Admin

@eunpalim will present next meeting (06/07/23), @ars599 the following meeting (20/06/07)

See speaker spreadsheet for details:

Tilo will be back in Australia in August, will consult with community on moving meeting to a new time slot.

Resource use

@tiloz has used 0.7MSU, will use 1.25MSU by the end.

@ars599 can use up some of the spare resources on CM2 SSP ensemble runs.

Make time next meeting to discuss shared experiments.

ACCESS-NRI Workshop

@wilma attended last meeting. Workshop Sept 5-8. First day is training day. Tue/Wed main NRI Workshop. Thu/Fri for working groups.

Monday training is mostly NRI organised: general skills, linux, git, GitHub, gadi. How to run a model, evaluate model output. Idea was to CM2. Too specific might not be a broad interest. Asking for input, let whoever is attending meeting know, can report back. Wanted to cover model evaluation, but might be too specific, e.g. iLamb.

Potential to have ESM WG specific training. COSIMA is running a hackathon. WG could do a changing land/sea mask-athon.

Next meeting to discuss program for NRI Workshop days, Tues/Wed. Input welcome.

Last meeting decided we only need one day. Need a committee to plan this? Adele is organising COSIMA meeting. Might be some overlap. Could either join, or join for some sessions/social events.

Fill in the poll!

Working group meetings will be in parallel. Affects this group more than others. Parallel and joint sessions? Need to think how to organise.

Should at least have a coupled modelling meeting in the morning of Thursday.

There will be a registration form for attending the workshop, will have the ability to indicate a desire to present a poster and/or a talk. Will go out shortly (next week?).

Monday training is still being decided. If ESM community members have training they want to deliver, please contact @rogeredberg. NRI can assist and facilitate this.

@Dietmar_Dommenget would like to hear about the model development plans.

Ben Evans has committed to having some NCI training.

@dkhutch will attend the ACCESS-NRI Workshop Program Committee meeting on behalf of the ESM WG.

CMIP7 Roadmap

Talk by @RachelLaw (CSIRO):

Slides: ACCESS-development-210623.pptx (481.4 KB)

CMIP7 waiting on IPCC meetings to fine-tune their timing.

FY 23/24: Model dev
FY 24/25: Model testing
FY 25/26: Base model simulations
FY 26/27: Additional model simulations
FY 27/28: Maximising impact

Australian CMIP7 Consortium: CSIRO (lead), ACCESS-NRI, NCI, Bureau of Meteorology, Universities. Have approached department. Can continue in the meantime without specific funding. Will require an oversight committee, project leadership team and project coordinator/manager or similar.

Model development paths: no one route to a model configuration for CMIP7.

Department wanted “options”: have an ambitious, with some less ambitious incremental possibilities.

Feedback:

  • Definitely need a “fast” ESM model. Perhaps need a definition of “fast”. ESM1.5 is 60 MY/day. Would like something even faster, Dietmar had 100 MY/day simulation. Would like this to be a possible configuration.
  • No separate group for model development. Perhaps in the future when the consortium is spun up there will be separate meetings.
  • Will model dev will be parallel paths. Lots of simultaneous development. At some point decisions will have to be made. Start with ambitious, target CM3/ESM3 but must be realistic and make sure we have something viable.
  • Don’t know timing of CMIP in detail.
  • Want to use same model code, but switch options to enable slow/fast runs.
  • Global stocktake will be in 2028, and that won’t change. There will be simulations required for this, and need to be required by 2027, regardless of CMIP timelines. Might need “targeted simulations”.
  • A lot of discussion about emissions driven simulation, strong support from community to make this the standard for Scenario MIP. Need full interactive carbon cycle. Doesn’t exclude concentration driven, but focus will be emissions driven. Any net-zero simulation requires emissions driven.
  • Paris Agreement and government targets are around emissions.
  • Stocktake is about countries meeting their stated emissions targets.
  • Is model development technically or scientifically driven? CABLE dev is driven by land use change for carbon accounting. Also aspects of ENSO are not so good in CM2 and need improvement.
  • Once there is a model with output then community input can be incorporated.
  • Model evaluation key to this to catch issues in development pathway.
  • It may be good to have some talks in this group on what we think could be improved in the ACCESS climate model

What’s going on?

Any plans to improve WOMBAT? @matthew.chamberlain: Yes, Pearce will be starting later this year. Work on getting nitrogen/phosphate cycle into WOMBAT.

Any plans to use TOPAZ or COBALT from GFDL? WOMBAT is what we’re used to using. BLING is in MOM6 but very basic. COBALT is in MOM6. WOMBAT is intermediate in complexity between them. Need advice from BGC community about what to use. MOM6 can have long tracer time steps, maybe by a factor of 10x. No-one actively working on this in MOM6 in Australia. Would like to do so.

Strong zonal current along equator in CM2 0.25, @ars599 might have found the issue, and if so could run another 400 year run and share with community.

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me

6

Date: 06/07/2023

Participants: 13

Chair: @LaurieM

Admin

Change next meeting to 3.30pm. Science talk volunteers.

Working Group Experiments

@LaurieM Some paleo experiment used across community? Mid-pliocene experiment? Good to have experiment that could benefit a number of communities.

@Aidan COSIMA example of control runs at various resolutions with repeat year and interannual forcing. Used as base for perturbation runs.

@lauriem High resolution coupled model for sensitivity experiments?

Goal is to do something that would not be done otherwise, but has some example projects.

@Dietmar_Dommenget would like to run a lower resolution coupled model. Don’t have current ability to do so. Change land/sea masks. Good for El Niño and paleo. Might be worthwhile group project to get going. Don’t have resources to do this alone. Requires human time rather than compute time.

@LaurieM More like Hackathon? Could be training day activity? Might be a decent start.

@dkhutch has similar problems. Currently stuck on land/sea mask modification. Would be interested in a hackathon or similar.

@dhb599 not trivial work. Looking forward to this assistance as well from ACCESS-NRI for high resolution ocean coupled model. This work will be more common in the future. Would be good for whole community.

@LaurieM working group meeting on 17th of August to focus on this? Invite Davide and Martin to get some movement on this issue. @dhb599 they know what needs to be done. Automatic package would be great. Don’t know progress. How far they have gone, what to expect?

@dhb599 There is a CM2 version with a wave model. Done some experiments. Could run some experiments of interest to community. Would need more resources, compute and data.

@ClaireT COWCLIP - coordinated ocean wave climate intercomparison project. Needs >= 3hr surface winds in particular. Prelim coupling of WW3 to either ocean or atmos model has been accomplished but I’m not sure anyone’s run it in a full ESM yet (maybe UKMO?).

ACCESS-NRI Workshop

@dkhutch gave overview of workshop draft. ESM community need to have some speakers if want to get ESM in workshop.

@RachelLaw Tilo was approached if interested in another ESM training session. Maybe if we want other ideas around modifying experiments, or other ideas. Should contact @tiloz with ideas.

Poll for training on Monday? Can decide what is appropriate. How to run.

Follow up with ideas for training on forum: Ideas for Workshop training?

Need everyone to answer the poll

Science talk

Eun-Pa Lim “The impact of ozone forcing on the 2020 Antarctic super vortex and associated surface climate conditions”

  • 2020 super vortex in Oct-Dec was an unusual event.
  • 2020 ozone anomalies significantly cooled the Antarctic polar cap in the lower stratosphere and strengthened polar vortex. Ozone forcing needs to be improved in our seasonal forecasting systems.
  • ACCESS-S2 is predicting the 4th consecutive vortex strengthening event in the coming spring.

ACCESS-S2 still using climatological ozone forcing. @sofarrell up to date ozone data was difficult to put in code which uses zonally averaged regions. GC5 uses full 3D ozone fields.

@Dietmar_Dommenget Why is the current forecast more reliable than previous inaccurate forecast? What is the driver of this event? Previous bad forecast was ACCESS-S1. Now upgraded to ACCESS-S2. The 2020 forecast deficiency was unusual. Hindcasts were good. Spring Polar Vortex are usually predictable due to well understood process. Already an anomaly in the stratosphere. Current ozone anomaly lowest on record. Probably related to unprecedented stratospheric water vapour due to volcanic eruptions.

What’s going on?

@paulleopardi started at ACCESS-NRI this week. Finishing off auditing of ancillaries in /g/data/access project. Will be involved in ancillary work in the future.

@Aidan will put together some workshop related information together and send and update to the Working Group announce topic to make sure everyone has a chance to have their input.

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me

7

Date: 20/07/2023

Participants: 18

Chair: @ShayneM

Admin

Volunteers for science talks welcome. If have talk relevant to group please let us know.

@dkhutch has volunteered for 31st August.

Interested in very relevant/important watching a pre-recorded video for a speaker? Be good if it was something that would spark discussion in the group. Need someone to sponsor a talk. @eunpalim saw a talk about high resolution impact on ENSO and southern ocean processes.

Definitely not a regular event. Mostly about filling gaps.

Meeting structure discussion: Working Group seems to divide along between different model development aims:

  1. improving processes and their representation in models
  2. improving performance.

Is it worthwhile to alternate focus of meetings between these two aims to ensure community adequately catered for?

@Dietmar_Dommenget not sure if there is a 50/50 split in the community. Everyone wants to improve models, once have a fast model not such a focus. Can improve model by increasing resolution, or by improving parameterisation. Don’t have to always be interested in specific equations.

@RachelLaw Perhaps split is around the models. A lot to be gained by sharing tools and workflows around models.

Shared experiments/ACCESS-NRI resources

Any proposals for experiments to utilise ACCESS-NRI compute resources?

See the topic for more info.

@Dietmar_Dommenget if can get a modified land-sea mask can use resources. Make for @dkhutch. Relies on getting a breakthrough with configuring the model.

Would be good to have a board displaying all the experiments, and how they are done.

@dkhutch has made a wiki with instructions for making modifications to the ocean model in GFDL CM2.1.

ACCESS-NRI Workshop

Draft program now on ACCESS-NRI website

Breakout discussion should be relevant to more than one area (Working Group).

There is an online spreadsheet where can suggest speakers. There is a slot for a speaker from the ESM Working Group.

Talked to @gpontes, @Dietmar_Dommenget and some others, that would be good to have an hour in the Working Group day toward model configuration modifications, land-sea mask, boundary conditions.

There is some space late additions to working group day if results are available.

Pencil in talk for faster ACCESS Coupled model by @Dietmar_Dommenget. Get paleo group interest.
@dkhutch would also like to present something model development/configuration related with some UNSW colleagues.

Need to feedback to @wghuneke as she is organising the working group day.

@paulleopardi Bureau still pursuing coupled models, maybe get their experience with using Nemo, could be interesting interchange of ideas.

Science talk

@ars599: “The representation of ENSO in ACCESS CM2 and HadGEM3”

Work from his PhD research.

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me

8

Date: 3/08/2023

Participants: 21

Chair: @eunpalim

Admin

Volunteers for science talks welcome. If have talk relevant to group please let us know.

Francois Delage is happy to present in October. No-one in available in September.

Changing meeting time permanently to 1pm? Can make a fixed calendar with reminders.

Over half voted yes. Settle on 1pm. Open up discussion on forum. Perhaps discuss at working group day.

ACCESS-NRI Workshop

Draft program now on ACCESS-NRI website

  • Does WG want ACCESS-NRI to organise an eventbrite event for the working group day?

Yes. @Aidan will follow up with @wghuneke

  • Natalia wanted to have all suggestions for invited speakers by Wednesday 26th July. Several suggestions were put forward by ESM working group in time.

  • We have not nominated a speaker to represent the ESM working group and give an update (10 minutes I believe). At the moment Tilo Ziehn’s name is in the spreadsheet with a question mark. This should be discussed / confirmed or not at the ESM working group meeting.

Tilo happy to give presentation representing ESM Working Group.

  • Natalia needed requirements for each working group’s workshop to be submitted ASAP. ESM group has currently requested half a day (Thursday morning). Can we please confirm with the ESM group whether half a day is what we want? (E.g. do we want a whole day?)

  • Other requirements they need are:

    • catering: How many people? (this is the most time-critical for organising, currently we say 10-20.)
    • presenting resources: Do we need posters? (I think yes) Do we need video conferencing? (not sure)

  • A general request to please get our registrations done for the ESM workshop as soon as we can. They have a registration form template if we need it: PROGRAM Template (for WG meetings) - Google Docs

  • Reminder that the deadline for the workshop registration is 15th August. Oral presentation deadline already passed (31 July), but I think posters are still able to be submitted (and automatically accepted).

No-one has volunteered to speak at Working Group Meeting. Perhaps meet to discuss shared model experiments that can be used for perturbations.

Have a posters, can talk to each other. Don’t force people to talk who don’t want to.

@eunpalim not sure if focus on WG day was science, technical or getting to know each other. So didn’t volunteer in case it wasn’t appropriate for what we were doing.

@RachelLaw Perhaps people who volunteered and can’t talk at main program can talk in working group.

@wghuneke Eventbrite yes, but can’t really afford to wait for an Eventbrite or another poll.

@JulieA would be good if the ESM group thought about CMIP7, and what experiments we need which would inform what model(s) we need.

WG have requested poster boards. Can re-use posters from main session. Moving to discussion about experiments we want to do as a group. Make a time and go with this.

@Dietmar_Dommenget How do we develop future versions of ACCESS model? Visions. Better El-Nino representation, and who would do this work?

There are breakout sessions that might overlap.

@ShayneM Could we talk about what the current models do well and poorly compared to other models. Maybe even previous versions. Are we progessing? Model evaluation for various topics we’re interested in? This could have been done already, would be good to know.

@Harun_Rashid there is a good paper that describes both access models and comparison with cmip6 models.

@ars599 Could we have a fixed agenda item with a single slide with current model evaluation?

@wghuneke have some ideas. Will take this offline, and discuss with a smaller group and propose an agenda for the working group day.

Seem to have some keys points for the WG day: model eval, update on how model is performing, model experiment directions and bring posters.

Science talk

@wghuneke “ACCESS-CM2-025: CM2 with a higher resolution ocean.”

  • Increase in ocean resolution results in similar patterns in SST bias, but even larger and worse in southern ocean
  • Similar with salinity
  • Warming surface air temperature with increased resolution
  • Surface speed has more detail, more than doubling of kinetic energy going to higher resolution. Coupled model overall higher values compared to OM2 model
  • Max mixed layer depth: still getting open ocean convection at higher resolution. Disappear in 0.1 model where they appear on the shelf. Labrador Sea convection missing in quarter degree, but visible in OM2 version
  • Seal level anomaly standard deviation is much better than low resolution. Coupled performs better than ocean only
  • Stronger winds and evaporation in higher resolution
  • Coupled model displays multidecadal variability that is missing in lower resolution and ocean-only models
  • Multidecadal is visible in upper ocean as well as AMOC
  • Source of variability is convection in the Labrador Sea when it is ice-free
  • This source of variability is not observed. Not unique to ACCESS model
  • Future plans
    • Updated CM2-025
    • Develop CM2-01
  • Current testing OM2-01 with CM2-025 atmosphere

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me

9

Date: 17/08/2023

Participants: 16

Chair: @tiloz

Admin

Speakers: Francois Delage can speak on September 28th: “CMIP6 attribution of extremes”. @lynnzhou working on attribution of climate extremes (S2 model) and can talk on 26th October.

Anyone who wants to change the meeting time should comment on the forum topic:

Shared experiments

@Dietmar_Dommenget can’t manage to make the necessary configuration changes to be able to run proposed experiment. Uses “fast” version of N48 (30 years/day). Some tracers have been turned off to make model faster. Might be a changed vertical resolution from CM2.

@eunpalim proposed ACCESS-S2 Experiment to understand influence of atmosphere on last triple La Niña. Don’t necessarily need compute, but storage support required. 1 member takes 1TB. Joint investigation with F&P Working Group . How much interest does ESM WG have for this topic? If compute resource unused, would use some. Minimum requirement would be storage, perhaps shared with Forecasting and Prediction. For 20 ensemble members would require 20TB. A lot of interesting atmospheric events between 2000 and 2022. Related to ozone depletion. And volcanic eruption. There is an idea Hunga Tonga–Hunga Haʻapai eruption allowed La Niña to persist. This experiment is designed to test this. Could start running immediately.

Community members who are interested in this experiment should reply to the experiment proposal to discuss the diagnostics they would need to collaborate and achieve their science goals.

@tiloz WG can request storage for interim until funding stabilised. Likely long-term have 100TB available.

@LaurieM need to answer to @clairecarouge with our requirements. Might need a procedure and policy for for retention of data in the longer term.

ACCESS-NRI Workshop

  • Invited speakers are now mostly locked in
  • Big focus is now working out how the breakouts will be run
  • Sitting at about 100 in person attendees at the moment, registrations will all be extended to midnight 20-Aug (BUT we’re almost at capacity for in person, so register quick if you haven’t)

Program:

2-day workshop will be hybrid, working groups will be limited.

Thursday 7th September ESM Working Group meeting. Half day to join other WG meetings.

Will be two sessions. First session with presentations focussed on model evaluation and development. Can indicate on registration form if you want to present.

Second session will be posters, can talk about anything. Learn about what we do and how we can collaborate.

Registration open until 30th.

Science talk

Romain Beucher, Model Evaluation and Diagnostics Team Lead: “ACCESS-NRI support for Model Evaluation and Diagnostics".

Download PDF version of Romain’s slides (1.5 MB)

Will be presenting at ACCESS-NRI Workshop. Also running training on the Workshop training day, and will run other trainings in the months afterwards.

  • Will be supporting a number of tools, e.g. iLamb, ESMValTool
  • Aspire to include analyses from ACCESS Community
  • Created “ACCESS-NRI Replicated Datasets for Climate Model Evaluation”: observational data collection at NCI (ct11).
  • Created ACCESS Intake catalogue for data at NCI. Dougie Squire will be giving some training on the Workshop training day

Q&A

ACCESS observational data collection is to support the tools. Yes some data already available at NCI, will also be making sure NCI datasets are up to date and in good condition. Some datasets on NCI might be moved or replicated.

Would be good to state if data are original, or interpolated to specific grids.

How are model components prioritised? Tool focussed: ESMValTool to begin with. Mostly ESM and CMIP. Assume CMORised data. Also trying to make tools aware of native model output to CMORise on the fly.

What is monitoring on the fly? Working on a tool to evaluate model output as the model is running. Currently doing basic comparisons. Generating simple plots on the fly. Generalisation of analysis that @MartinDix has done in the past.

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me

10

Date: 28/09/2023

Participants: 18

Chair: @eunpalim

Admin

@dkhutch could do a talk in November. @jemmajeffree could also do a talk in November. CLEX Workshop 20-24 of November, so need to check if we have a meeting on 23rd.

Shared experiments

@eunpalim when is next SAC meeting? Need approval for ACCESS-S experiment.

@gpontes asked about total compute: 875KSU / quarter, maybe more if others can’t utilise all the time, and this might still be revised upwards depending on NCRIS funding announcements.

50TB of /g/data space for the ESM Working group. Hard to know when this will be available. Again depends on NCRIS funding announcements.

@Dietmar_Dommenget priority should be focus on ideas that push ACCESS models forward. But lacking expertise to develop models. Without support to develop there is no need for compute time and or disk space. Researchers don’t have technical capability. Land-sea mask changes still not possible. Took one week with GFDL model.

@dkhutch still can’t manage large changes.

Needs an engineer to sit with a researcher, fix their issue and take this approach and make it general.

@Aidan to progress this internally. @dkhutch happy to write a proposal document outlining what is required, what has been achieved and what still needs to be done.

Working Group Day

Thanks to @wghuneke for organising, and @dkhutch for participating on Workshop planning.

Everyone can please read the notes. Reply with thoughts on action items:

Maybe if other working groups have main meetings outside of workshop, it might be a good time for ESM WG as everyone is already there.

@eunpalim successful just from a networking experience, got to meet everyone and know who everyone is. Very enjoyable.

Working group photos are here:

Forum overview and questions

@Aidan gave a quick overview of using the forum and emphasised that everyone should join the ESM Working Group.

Could benefit from some categories for each model setup as they use quite different infrastructure. Would benefit from a category for ESM and CM. Maybe also worth having a Paleoclimate category.

What’s going on?

Thank you for submitting abstracts for AMOS. Should we have an ESM Working Group dinner?

@HIMADRI_SAINI has an experiment would like to run. Maybe not interesting for the whole community. Would be happy to burn unused time. Guidelines for use are available on the forum

@wghuneke planning on running CM2-025 with changed bathymetry and some other fixes. 500 year present day control run planned. Will do a pre-industrial control run next year. 12.5KSU/yr. Still using eddy parameterisation in ocean model. @sofarrell NEMO uses latitude dependent Halberg parameterisation HadGEM3-GC5 uses this in its 1/4 degree version and its available in MOM6.

@gpontes Discussed control runs during working group day. PI vs present day. @wghuneke interested in both: present day can compare to obs to introduce model. PI to do sensitivity experiments. @gpontes observations may already be under some transient forcing. Comparison with present day runs may not be a clear comparison.

@sofarrell tend to use PI at CSIRO. Only in the last year or so has present-day controls been used more often. UKMO tend to tune up with present day. PI controls have a very strong cold start in first few hundred years. Took 500 years to reach starting global temperature. Maybe due to top of atmosphere and deep ocean heat uptake. Also made changes during early part of CM2 runs, and used Jules initially before switching to CABLE. ACCESS1-0/1-3 faster to spin up, ACCESS ESM PI-control came from much longer spin up as the carbon pools needed to be stabilised as well. The published CMIP6CM2 and ESM PI controls start after they have stabilized.

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me

11

Date: 12/10/2023

Participants: 16

Chair: @LaurieM

Admin

First meeting in November overlaps with ACEAS.

New sub-categories on the forum. To get notifications of topics in the sub-categories at this stage you must watch each sub-category separately.

Put ideas for working group meeting, including suggested speakers, should go in this topic:

Shared Experiments

@LaurieM: Want people to use it. Needs to be fair. There are instructions on how to put forward an experiment to use WG resources:

@LaurieM: Ideally if this is done early enough can discuss and decide between proposals. If done late, becomes first-come first-service. Currently lagging putting projects together.

Three experiments used WG resources, two of them were described here

@ars599: Would be good to have other people than just ACCESS-CM2 using this time.

Science talk

@lynnzhou (Bureau) “ACCESS-MICAS: a forecast-based tool for event attribution”

What’s going on?

@dkhutch have a Miocene test case running. Significant step forward.

See related topics:

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me

12

Date: 26/10/2023

Participants: 19

Chair: @tiloz

Admin

@tiloz: ESM Working Group clashes: ACEAS Meeting on November 9th, CLEX 23rd. Please answer poll for meeting times to assist with deciding if the meetings should go ahead.

Working Group Resources

Shared Experiments

New NCRIS funding. 50TB for lg87. Currently no community proposals for how to use that storage.

@tiloz: Still no proposals for shared experiments for this quarter. Can use lg87 storage for shared experiments. Usually struggle for storage.

@eunpalim: Waiting for SAC to approve proposed ACCESS-S2 experiment. @tiloz will mention as part of Working Group update to SAC today. Minimum requirement was storage as join project with F&P WG. 10TB from each. Would be happy to utilise computing resources.

@LaurieM: Interest in running a standardised CMIP6 future scenario with higher resolution model @wghuneke is working on?

@wghuneke: Can run an experiment if required. Maybe defer to @eunpalim to run first. Currently running 0.25 deg present day and plan to do pre-industrial. For the standard (no BGC) CM2-025: cost 12.5 kSU/yr and walltime ~4 yr/d

@sofarrell definitely want to be able to share runs. Had some issues around Antarctica but that should be fixed.

@dhb599 Interested in new runs. Want to run with ocean-bgc. Previous run wasn’t clean due to crashes. Will run on CSIRO project. Important delivery for NESP project, as a base for ACCESS-CM2 0.25 run.

@tiloz Model crashes affect reproducibility? Long enough period would still be useful to the community. Use for initial analysis. @dhb599 can make available. Ran over 275 years with ocean bgc. Be careful using it because crashes means we can’t reproduce it. Runs at less than 3 model years/day. Not sure about compute and storage requirement.

@dhb599 need help to scale ocean. Maybe need some COSIMA help. Could maybe run 4 years/day. So no work done on efficiency and tuning. BGC adds a lot of time and resource use. Need to talk to @matthew.chamberlain

@AndyHoggANU ocean will scale, should be able to match the atmosphere.

@wghuneke Runs through accessdev rose/cylc on accessdev. About to start running on ARE on persistent session.

@AndyHoggANU Would it be useful for ACCESS-NRI (@paulleopardi) to do some model load-balancing optimisation? Answer: Yes!

@tiloz would be great to have faster throughput, 5-6 yr/day vs 2-3. Also useful for future CMIP7 work, but can benefit immediately.

@AndyHoggANU mentioned that new long Rose/Cylc based experiments should be run from persistent sessions rather than Accessdev.

@Aidan need to put shared information on the forum rather than in emails. Aidan will make important topics more visible.

ACCESS-NRI Update

@AndyHoggANU ACCESS-NRI has been slow to release compute time and storage because agreement ran out in June. Have been waiting for NCRIS funding announcement. Now been announced. Received most of requested funding.

Total data allocation is 3-4PB over 4 years. This includes data collections, working group storage and ACCESS-NRI storage.

If there is a run that is of interest for whole working group can keep it in /g/data/lg87. Can index, make it easy to access. If there is sufficient meta-data, can sustain and manage the storage.

Building up expertise with ESM2 type model is critical for the community.

If there is something the WG chairs think will benefit for community, ok to say yes. For @eunpalim’s experiment collaboration with another working group is fine, but better if all under one project. ESM is probably the better fit.

Major aspect funded: efforts to support CMIP7 and advances in CABLE4 LSM. Asked for 4 years of ($5M) gave us $2.5M. Interpreted as funded for 2 years. Re-apply for 2 more years. Feeling is NCRIS want to see how CMIP7 consortium comes together.

Will be hiring more people in Claire’s land team. Also in model evaluation, atmosphere, ocean and coupling. Will have to be fixed term appointments. Will help CMIP7 development effort. Also help to fund people working on CMIP7 who are currently funded off recurrent funding.

Have a current recruitment round just closed. Will have another early next tear.

Science talk

Cancelled.

What’s going on?

Changing coupled model land-sea mask

@paulleopardi What has been happening for ancillary generations for paleo runs?

@dkhutch recently got Miocene land-sea mask working in ACCESS-ESM-1.5. Have a number of scripts that are accessible/readable but will make available.

@paulleopardi want to know how it relates to other work, particularly what @atteggiani has been working on. Might be some possibility for rationalisation.

@dkhutch used older CMIP5 generation tools. Worked for me.

@tiloz understand there is a more long-term fix, but we do need to be able to work in the mean-time. Davide has been very helpful. Need quick fixes but do need a well documented tool.

@dkhutch is your work related to the request that Dietmar had for N48 land-sea mask? It’s related. Have had some discussions. No entirely sure. He is using rose-cylc, some of the interpolation steps are being auto-generated, and might be going wrong. Will give him the routine that worked and see if that will work for his problem.

@tiloz want to make sure people stay connected. Would not be a good outcome to let it drop. Small changes work, larger scale falls over. Similar to my experience.

@dhb599 reconfiguring models is critical part of workflow. Looking forward to this work. Want to change CM2 and easily land-sea mask and change related. Want to see that script working for everyone, not just specific case.

@tiloz want flexibility to support multiple resolutions and configs. @dkhutch great to know others are interested.

Issues with different run systems

@LaurieM followed issues on forum. Due to different systems (script/payu/rose-cylc) there are issues with information exchange. We want to run ssp2.6 run. Currently not possible with forcing. Could run 8.5 no problem. 2.6 has a problem.

Emissions driven scenarios

@tiloz: strong focus on emissions driven simulations in CMIP7. Trying to prepare for that through new international multi-model experiment. Some experiments we can run with ACCESS CMIP6 version already. The CMIP6 DECK includes the 1% concentration driven CO2 increase run which is the basis for estimating TCRE and for other experiments around ZEC and reversibility as well. We need equivalent emissions driven run with constant emissions (e.g. 10 PgC/yr). This will become part of CMIP7 and will allow for a better estimation of TCRE and basis for ZEC and reversibility assessment without switching between concentration and emissions driven mode. Currently running ACCESS-ESM1.5 with constant emissions over 100 year period and then switch to zero emissions. Happy to share results. (Thanks to @atteggiani for helping to generate the required CO2 emissions ancillary files)

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me rather than replying.

1 Like
13

Date: 23/11/2023

Participants: 11

Chair: @LaurieM

Admin

  • Plan to have one just more meeting this year (Dec 7th) and no meetings in January. First meeting of 2023 will be February 1st.
  • We have two volunteers for December and February (thanks @jemmajeffree and Francois Delage). Would be good to have at least one more speaker for mid February locked in before break.

Shared experiments/ACCESS-NRI resources

  • ESM WG project lg87 has /g/data allocation of 50TB, same for /scratch and F&P WG transferred their compute allocation for this quarter, so there is 1.75MSU total.
  • @eunpalim’s Triple La-Nina experiment: @griff and @lynzhou running this. Done 1 start date so far and used 130KSU/1TB so far. Will run 17 more.
  • Any proposals for shared experiments for next quarter? See guidelines for how to do this. First meeting is in February, so not much time to organise to utilise resources.
  • ACCESS-NRI Merit Allocations Guidelines: Release scheduled for today. Overview from @Aidan and discussion/questions.

Science talk

Works for Miocene, but still haven’t problems with orography. 7 orography variables in the UM restart file. Works with small changes, but trying to just change the areas where miocene changes leads to a model crash. Can’t find how this was originally generated.

Wrote a user story on the hive

all code available on github:

What’s going on?

ACCESS-CM2 updates

@dhb599 ACCESS-CM2 PPE Experiments using UKCP18 PPE Settings. ENSO not good. Quasi-biennial life-cycle of ENSO. 7 out of 10 metrics are good. Only ENSO stands out as bad. Run group of 8 PPEs. Large TOA Energy imbalance: 0.4 W/m2.

Chose best and ran more tests. Less than 0.2 W/m2 vs 0.4.

Experiment proposal

@gpontes SSP585 with meltwater experiment proposal

ACCESS Workshop

Videos from the workshop talks are now available on the Forum if you missed a talk, or want to watch again.

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me rather than replying.

14

Date: 7/12/2023

Participants: 13

Chair: @tiloz

Admin

  • Last meeting of the year, next meeting is February 1st 2024
  • We have a speaker for the first meeting of next year (Francois Delage)
  • @JulieA and @RachelLaw will give a CMIP7 update in mid-February. Will have timelines from IPCC at that stage. Bureau meeting mid January. Have a proposal for fast-track experiments. Could give an update by then. Could have a discussion in this group.
  • @eunpalim Any AMOS attendees could give their talk to group. @tiloz yes, but doesn’t have to be a polished talk.
  • @JulieA ACCESS-ESM WG will organise a meet-up at AMOS. Could be lunch or dinner after combined session with Paleo modelling.

Shared experiments/ACCESS-NRI resources

  • ACCESS-NRI merit allocation guidelines released
  • Have proposals for experiments for next quarter from @wghuneke (storage) and @tammasloughran.
  • @dkhutch will put up a proposal for next quarter
  • @AndyHoggANU Don’t currently have tape storage for Working Groups. Wasn’t clear this was wanted.
  • @gpontes will we have more /g/data storage space in future? @tiloz perhaps, but do need a strategy for migrating to longer term tape storage. @MartinDix there is a ramping up schedule from June next year, but NCI don’t have disk storage available. Might not be more for next 6 months.
  • @tiloz would be good to have a discussion topic on the forum to help inform how to make management decisions.
  • ESM WG Chairs will give decision on @wghuneke storage request before Christmas break.

AMOC Freshwater forcing experiment

@gpontes

  • Experiment adding fresh water forcing to capture AMOC behaviour for past century. Current models are not able to capture observed AMOC behaviour. These experiments can capture observed AMOC, now extending into future projections (ssp585) and designed freshwater forcing that should be consistent. Doing a 4 member ensemble out to 2100 will finish this year. If results are promising and add ensemble behaviour. Capturing AMOC behaviour makes experiments potentially interesting for this group.
  • Currently post-processing with ACCESS-Archiver. Not doing CMORisation.

Role of Atmosphere in triple La Nina

@eunpalim

  • Initialise ACCESS-S2 with realistic ocean, and random land and atmosphere
  • Want to make data available for other researchers
  • Will post-process and keep only the required data. If others want specific fields/variables let @eunpalim know.

Science talk

@jemmajeffree “Using model analogues to determine the impact of various ocean basins on ENSO”

What’s going on?

  • @MartinDix and @dkhutch will talk offline about ACCESS-ESM-1.5
  • @RachelLaw CMIP7 Announce topic. Watch this for updates. Separate topic for commenting on fast track v1 proposal. Will submit ACCESS feedback next week. There is one paleo experiment suggested for fast-track: an abrupt change from a PIControl. Would seem do-able, but would value paleo modellers feedback about the possibility.
  • CMIP panel have drop-in sessions, recordings are available. MetOffice only doing emissions driven experiments. Trying to decide on DECK, and many modelling centres only want emissions driven experiments. @RachelLaw given DECK required historical and ESM historical might as well have emissions and concentration driven runs, but weight emissions-driven. Looking at revising DECK experiments based on feedback.

This is a wiki post , if you want to update any of the details in these meeting notes, or to add your own recollection of what was discussed then edit me rather than replying.