This topic is to collect ideas from the community for discussions at the ESM Working Group meetings.
Please reply with your ideas. Consider adding a bit of context if that will help others understand your idea.
- How to modify UM ancillary files (moving the continents) for deep time paleoclimate simulations.
- How to modify land model boundary conditions for deep time paleoclimate simulations.
- How to configure OASIS coupler for novel ocean and atmosphere topographies and land-sea masks
- How to streamline ACCESS model outputs into compressed netCDF4 format and save storage space.
3 Likes
@atteggiani is working on a library to manipulate ancillary files and there is a topic to collect known ancillaries:
You should definitely add anything there that you need to manipulate for paleo work, so that use case is covered.
These first three are technical questions. Do you want to have a discussion about how people do this currently, in case there are already good approaches you don’t know about?
I’d say this is a fairly generic requirement for most models. Are you using the APP4 post-processor?
In general it sounds like you’re asking for better pre- and post- processing tools for paleoclimate. Would that be correct?
2 Likes
Hi @Aidan, thanks a lot for your reply. Indeed I am aware that @atteggiani is working on the UM ancillary file preparation steps. I just wanted to make sure this issue was on the agenda… so we can discuss where we are up to on that front.
Yes it’s true that the first three are technical questions… the reason I specifically mention the land model and OASIS coupler separately is because I really don’t know who to ask about modifying those things.
Thanks for the link to the APP4 post-processor, that is very handy to know. I also heard from @chloemackallah via Nick Yeung that there is an ACCESS archiver tool GitHub - ACCESS-Hive/ACCESS-Archiver: ACCESS Archiver is designed to archive model output from ACCESS simulations. which helps to compress the outputs quite well. @gpontes has reported very good space savings from that tool.
Absolutely, that is very much a goal for me and others doing paleoclimate modelling at UNSW.
Sorry if it seems I’m being difficult, but I was just trying to get to the intent behind the queries. So I’d say an agenda item “Pre- and post- processing tools for ACCESS-ESM-1.5” would be a great discussion topic for an ESM Working Group meeting.
It would be good to have topics covering each of your questions, if there are not any currently, so there is somewhere people can post follow up information and questions. I’d suggest putting the ACCESS-ESM questions in the related category on the forum.
In some cases there is value in just documenting the current state, so create a user story, e.g. detailing a representative ACCESS-ESM1.5 output directory, the number and size of the files, and the same sizes after using the ACCESS Archiver. How that can be incorporated in your workflow in payu, that sort of thing.
By collating this information it provides a snap-shot for usage at a time, so you can check against that at some point in the future in case the composition of the files has changed, or the tools used aren’t behaving as you expect. Also others can compare their own experiments.
If you provide enough information to assess the effects of compression, for example, people with technical expertise who don’t use that model can give you feedback if it appears to be working correctly. Without this information they have no way of helping short of running the model themselves, which is a massive waste of their time, and so very unlikely to happen.
1 Like
I’d like to give an overview of how the forum works, what you can do with it, and why it is important for the community.
I’ve made a few short videos to help with using the forum effectively:
The next two ESM Working Group meetings clash with ACEAS (9th) and CLEX Workshop (23rd).
To help the committee decide which meetings should go ahead can you please indicate which meetings you can attend. Thanks!
- I can attend November 9th
- I can attend November 23rd
- I cannot attend either
0
voters
Sorry for the late notice, but there is an ESMValtool/CMIP7 Hackathon on this week which might mean many can’t attend a meeting at the normal time this week.
Could you please indicate below if you can make an ESM WG Meeting at 1pm on Thursday 14th (2024-03-14T02:00:00Z):
- Yes, I can make it
- No, I can’t make that date
0
voters
If we don’t have a decent number who can make this week we will cancel, and the next meeting would then be the March 28th.
Thanks!
Hi all!
@Harun_Rashid and I have been keen to kick off some discussions between users of ACCESS (specifically evaluation work) and the model developers. We were hoping to begin some discussions at a future meeting? We’d love to hear your ideas/feedback. Thanks!
Some context:
Dear colleagues and users of ACCESS,
We’d like to start a discussion among interested ACCESS users about how to effectively collaborate in order to improve ACCESS model performance. Many researchers do comprehensive evaluations of climate and earth system models, including ACCESS. They identify model biases, investigate the responsible mechanisms, and write papers on their findings. These results are useful to model developers and could potentially assist developers in identifying mechanisms and processes to focus on.
For example, what specific processes, or which aspects of parameterisation schemes could be responsible for the identified model biases?
We believe that bringing model evaluators and model developers together can facilitate discussions that’d be beneficial for both groups. Such discussions can even lead to some tangible model improvements. Also, model evaluators can gain some insights about the inner workings of the models, giving them the opportunity for a deeper interpretation of their results.
We propose to have such a discussion in the upcoming ACCESS-NRI Workshop, but it’d be good to get some conversations started before the workshop.
1 Like
Dear all,
I hope we can expand on this topic to have a broader title such as:
“Tuning the UM Atmosphere Model: Insights from PPE Experiments for Improving Weather and Climate Simulations”
I have some suggestions based on Christine and Harun’s excellent proposal for this session and the importance of understanding the UM atmospheric models and climate drivers on various time scales. So far in NRI, we don’t have anyone looking after the UM model except CSIRO, and we need help from the community. I know that in the new centre (W21C), Christine, Harun, and Jo Brown work on the Monsoon and ENSO, and Julie Arblaster focuses on the MJO and SAM, as well as cloud, land processes, sea ice, and other aspects. If we can attract more people, we might significantly benefit from their contributions to model development. Additionally, we might consider publishing an ACCESS PPE paper.
Recently, Dave Bi and I have worked heavily on the UM atmospheric parameters through Perturbed Parameter Ensemble (PPE) experiments. Initially, the idea was straightforward: we were looking for the “best” skewed ENSO, following Cai’s ENSO nonlinear theory. However, we found that these experiments might benefit the community.
This exercise will provide valuable perspectives on optimizing these simulations to improve our understanding and prediction of climate dynamics.
Regards,
Arnold
1 Like