If a mailing list does get created, I'd be v interested!

If you want project specific environments that are built into each project (eg for others to work off afterwards), I've found poetry solves a lot of the headaches conda creates around environment solving. Basically makes the environment part of the project itself and doesn't let you separate the two

Sorry yeah not super clear from me! My thoughts were that given observations have been trending towards La Nina while CMIP models generally trend towards El Nino over the recent past, would this imply that the model-estimated probability might be biased high?

Given ENSO seems necessary for such a jump, would it imply the probability of this kind of jump "should be" (ie. if the models reproduced the observed trend towards La Nina) even lower in climate models than calculated?

Really neat paper! Something I'm wondering after reading - how does the discrepancy between modelled and observed ENSO trends affect the qualitative conclusions?

Global climate highlights here: climate.copernicus.eu/global-clima...

I did something similar (but less well explained!) a while back and started to separate out the different components (at least into a few more groups: njleach.github.io/2020/12/29/H.... NB. this uses a different constrained FaIR parameter set & was based on emission-driven runs so some differences!

Thanks for setting this up, it's a great resource! Keen to be added if space. I'm a climate scientist at a climate risk analytics startup & postdoc at U of Oxford researching extreme weather attribution.