Hi all,
I’m currently trying to assess how plausible it is to run ACCESS-rAM3 from initial conditions that come from GCM simulations in the future - for example running ACCESS-rAM3 from an initial restart file of ESM in 2050 under SSP3-7.0. The model requires restart files in the UM fieldsfile/dump format.
I was wondering if these are available for ACCESS-ESM1.5 future runs (or other GCMs) under SSP scenarios.
Best regards,
Joel
Hi @Jatreutlein, that is something that I am currently working on (well was until I was pulled to a slightly different project). Please contact me through 21CW and we can talk about this with @mlipson.
Hi @Jatreutlein and @mlipson, I have been talking with you both via the 21st Century Weather private chat but I have also talked with @clairecarouge and we think it would be better for you to continue talking to the community here.
The original post was a bit confusing (in terms of what you want). We thought is might help to paraphrase the key thing you are asking for which is:
global ACCESS GCM scenarios start dumps and perhaps advice on how they can be run
The second part of this problem relates to enabling regional runs with that data.
Can anyone provide help on that front?
(Apologies for stopping the conversation earlier, I thought they were talking about the running of the regional model from CMIP inputs which is something I am working on).
i.e. a data access issue?
Have you asked CSIRO?
If they don’t have any restart files saved, ask the Met Office if there are any HadGEM restarts that could be shared?
What resolution global model restart is required? Worst case scenario could you run e.g. n96 ACCESS-ESM1.5 future runs yourself and archive the restarts? (Might need to do some work in the suite, my guess would be that housekeeping deletes them).
Ok thanks Bethan - I’ve asked the Bureau but not CSIRO yet, I’ll ask the Met Office too. Global model resolution currently is N216 but I could try with N96 if that’s all that’s available. I’ll have a look into your suggestion
@Jatreutlein the Bureau don’t run climate runs, that sits with CSIRO (or did 
). The question will be whether restarts were saved.
I think the person to ask at the Met Office would be Malcolm Roberts. Let me check with Julie. (The climate community is a bit outside of my network).
I only mention n96 because it’s cheap if you have to go to a last resort of re-running the model yourself! But I think you have compute (?), and n216 is the next least worst in terms of cost 
@Jatreutlein I just asked @eliocamp here at Monash as they have been running some experiments with the ACCESS ESM1.5 lately. They pointed me to /g/data/p73/archive/CMIP6/ACCESS-ESM1-5/$experiment/restart where $experiment contains many SSPs to choose between. I’m not sure how easy they would be to use for your particular case.
These use N96 resolution in the atmosphere. Note that no ACCESS ESM/CM global coupled model configurations have been developed (or run) with an atmosphere resolution greater than N96. The seasonal prediction system, ACCESS-S2, uses N216 but has a different ocean & land components and won’t have scenario restarts. Though Linjing Zhou, @Pandora and team have developed the MICAS attribution system that might be useful in your context.