Date: 29/9/2026
Chair: @dkhutch
Participants: 12
1. Admin and resource usage
-
Science Presentations: Spaces are available for presentations for upcoming ESM working group meetings. If you have anything you would be interested to present, or know anyone with a a topic that might be of interest, we encourage you to 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. People using lg87 resources for experiments are especially encouraged to present on their work.
-
LG87 Resource Usage:
- 109 KSU of lg87’s original 2MSU compute allocation has been used up. The remaining resources are likely to be used by the end of the quarter.
- 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. We strongly encourage you to submit proposals early in the quarter as it helps with resource allocation.
/g/data/lg87has increased to 82TB out of the 100 TB allocation. If you have data that’s no longer required, it would be much appreciated if you have time to clean it up!- If you are interested in using working group resources to run CMIP7 experiments, please get in contact with @tiloz Please see this topic for a list of the available experiments.
2. Journal of Southern Hemisphere Earth System Science special collection on Model evaluation for CMIP and IPCC AR7
- Please see this page for details on the special collection, which may be of interest for many in the ESM community. There will be a more in depth announcement on the upcoming special collection at a future working group meeting.
3. Science presentation: RiskMIP: Probabilistic climate projections for risk assessment
Jeremy Fyke and David Huard from the CCCma joined to present on the proposed RiskMIP modelling project. The team are currently seeking feedback on the proposal, and encourage you to get in contact if you have any thoughts or questions about the project.
Background
- Climate adaptation involves risk management and is greatly aided by probabilistic information about future risks. Earth system model projections produced during CMIP capture modelling uncertainty and uncertainty arising from natural variability, but not the uncertainty in future scenarios, making this risk quantification difficult.
- The RiskMIP proposal aims to provide planners and decision makers with probabilistic projections which incorporate all three of the above uncertainties, improving their ability to reason about climate risks.
Approaches
The RiskMIP team identified two approaches for producing probabilistic projections:
- The a priori approach: A probability distribution of future emissions scenarios is sampled from and used to force an ensemble of ESM projections. At minimum 4 model runs totalling 512 model years would be required.
- The * a posteriori* approach: Probabilities weightings are assigned to the 7 CMIP7 scenarios, and 3 additional simulations requiring a total of 384 model years are run to fill in the gaps.
Feedback:
-
Discussion with the ESM WG community covered topics including the value of providing providing distributions of projections compared to identifying a “most likely” scenario, the relative costs of the a priori and a posteriori approaches, the missing model uncertainty arising from shared biases, and what’s different between the role of RiskMIP and emulators.
-
Please see this post for further details on the RiskMIP project, including the RiskMIP concept proposal, and references on the proposed a priori and a posteriori approaches: RiskMIP concept proposal - Jeremy Fyke
Please feel free to correct any mistakes in these notes directly, or message @spencerwong with corrections.
Additional information
- Guidelines for accessing working group compute resources
- Share an experiment by starting a new topic in this sub-category.
- Join the ESM Group
- Subscribe to the ESM Working Group calendar:
- ESM Working Group Meeting Calendar feed (right click on the link and copy the URL)
- Use calendar feed URL to subscribe to the calendar