ESM Working Group day (7 Sep 2023)

[wghuneke (Wilma Huneke)]

Hi all,

You can now register for the ESM Working group day on Thu 7 Sep, 9am-1pm, in Canberra. The idea is to finally meet as a community in person*.

The half-day will be made up of two sessions:
Session 1: Presentations on model evaluation/development for the community to learn about what aspects of the climate ACCESS-ESM and ACCESS-CM simulate well/not well, and what developments are currently ongoing. The presentations will be followed by a discussion session, topic suggestions are given in the registration form.

Session 2: Poster session on any topic. We are hoping for many contributions to help each of us to understand the various research people are doing using ACCESS-ESM and ACCESS-CM.

*There will be an option to join online for session 1.

[wghuneke (Wilma Huneke)]

Thanks everyone who attended the ESM WG day last week. It was great to get together as a group and to talk about different aspects of climate modelling.

Below is an attempt to summarise the points that were raised during the day. Please feel free to clarify, elaborate, and add points.

CMIP7 / model development

• CM2 (or CM3) will be the main Australian contribution.
• Do we also want a fast model (ESM), e.g., for PMIP style experiments?
  ESM can be used as a development platform to update components (something like ESM1.6). Such a version would likely be used for science within Australia.
• What model improvements do we want most? Dynamical vegitation, BGC, …??
• Input by the community and especially new faces welcomed (i.e. people who haven’t been involved in previous CMIP rounds). Consider watching the CMIP7 Announce topic.
• Model evaluation hackathon: Get together as a community to work on model evaluation, including
  o teaching the community to use evaluation tools,
  o discuss variables of interest from a Southern Hemisphere point of view,
  o create recipes to do analyses.
  The idea is an outcome from the evaluation break-out of the NRI workshop, but also relevant to this group, get in contact with @nicolamaher for input/updates.

Model configurations – Work in progress/plans by individuals in our community

• ESM paleo configuration (@dkhutch)
• CM2-01, N96 atmosphere and 0.1° ocean (@wghuneke)
• CM2-N216, N216 atmosphere and 1° ocean (@nicolamaher )

If you’re potentially interested in these runs and want to give input on run requirements (output, Pi/PD/Historical/future scenario simulation, runtime length, etc), please get in contact with the respective person listed above).

Shared experiments

• @eunpalim is suggesting running an ACCESS-S simulation to study the role of the atmosphere of the recent triple La Nina

ESM WG Wishlist

• Transport matrices: provide scripts to (automatically) calculate a transport matrix for each model run. The output can be used for, e.g., BGC and is very quick to use once calculated. @matthew.chamberlain and @dkhutch have done this before.
• Workflow to use model output from coupled model to force individual model component.
  E.g. use atmospheric fluxes to run land or ocean model only.

[Aidan (Aidan Heerdegen, ACCESS-NRI Release Team Lead)]

All the talks have separate topics, which include the speakers slides:

Model analogues: Cheap forecasts and their implications

About From https://iopscience.iop.org/article/10.1088/2515-7620/acb90e Ding et al (2018) and (2019) proposed an alternative forecast method (Model-Analogs) to avoid additional forecast ensemble integration. In this method, the predictions are obtained from a long preindustrial control simulation by matching their selected variables to observed fields. The forecast members at various lead times are then taken from the subsequent evolution of these states in the control simulation. They showed s…

Evaluation of ENSO in ACCESS-ESM1.5

After a quick discussion with @Harun_Rashid regarding ENSO in a piControl experiment vs. an historical ensemble I did a quick comparison of the two. ACCESS-ESM1.5 is surprisingly stable. The equatorial Pacific SST is warming by ~0.5°C but all the rest seems to be exactly the same. @Arnold_Sullivan, @ShayneM may be interested. Feel free to share ths result with others. planton_evaluation_of_ENSO_in_ACCESS_pi_vs_hi.pdf (896.0 KB)

A Brief Summary of ACCESS-CM2 Development post CMIP6

Motivations Improve the model performance in climate simulation: Reduce model biases seen in the ACCESS-CM2 historical simulation for CMIP6 Comprehensive understanding of the ACCESS-CM2 ESNO simulation in comparison with other CMIP5/6 models around the world Efforts and Results Implementation of the FIO MASNUM Wave Model (MWM) into MOM5. Conduct PD simulations – for better comparison against observations. enhance vertical mixing deeper thermocline in the tropical ocean better ENSO lifecycle …

Last Interglacial subsurface warming on the Antarctic shelf triggered by reduced deep- ocean convection

About Enhanced meridional ocean heattransport and higher austral spring insolation let to reduced SO sea-ice cover. Fresher and warmer conditions meant higher stratification and weaker AABW formation and thus subsurface warming around Antarctica. Positive feedback paleo-ESM-Sep2023.pdf (10.7 MB)

If you have any follow-up questions, or wish to continue discussions please feel free to reply in the relevant topic.

[Aidan (Aidan Heerdegen, ACCESS-NRI Release Team Lead)]

@wghuneke

Model evaluation hackathon

Great idea! A Coupled Model Cookbook! The obvious approach would be to style it on the COSIMA Cookbook but utilise the ACCESS-NRI Intake Catalog as the primary data source, as it indexes so much more coupled model data.

@rbeucher might have some thoughts, as he has plans for a recipe portal, including upload and review mechanism (ping @max who is working on this).

So perhaps it isn’t as obvious to just copy COSIMA.

[Benoit (Benoit Pasquier)]

@wghuneke

ESM WG Wishlist

• Transport matrices: provide scripts to (automatically) calculate a transport matrix for each model run. The output can be used for, e.g., BGC and is very quick to use once calculated. @matthew.chamberlain and @dkhutch have done this before.

Sorry for unearthing this year-old topic, but I just wanted to advertise OceanTransportMatrixBuilder.jl, a Julia package intended for building transport matrices from standard CMIP output, in case anyone here is still interested in that item in the wishlist. This is still very early stage, but it should work for ACCESS models (ping @dkhutch and @matthew.chamberlain), and I thought it would be good to advertise it early on to get some eyes on it.

[dkhutch (David Hutchinson)]

Thanks @Benoit! Good timing as we have a new masters student @jaberg7777 who is starting to run some transport matrices to model oxygen in paleoclimate scenarios. He is starting with GFDL simulations of the Miocene, but is also interested in using ACCESS-ESM outputs. We will check out the Matrix Builder package, thank you!