Dedicated conda environment for model post-processing

[aidanheerdegen]

From #89 (comment)

conda/analysis3-25.09

None of the analysis environments are designed to be used in this way. They have no long term support, and the current plan is to remove them after a period of time.

The post-processing scripts should either be included in the payu conda run environment, or a dedicated environment should be created and deployed for this purpose.

[dougiesquire]

Yeah, the reason I tested this version was because I wanted to change to conda/analysis3. But it's a bit tricky at the moment because we've released a few versions that are borked. Fixing the version at something we know works felt like the lesser of two evils while we're still teething.

Note that it's easy for us to change this for all our configs if something does break because they all point to the same clone in vk83.

[dougiesquire]

Sorry, I didnt read your second paragraph before replying 🙈

The particular script in question is a stop-gap until intake-esm datastore generation is done by Payu.