Update use of fx vars to new syntax

[sloosvel]

Description

Changes in ESMValTool recipes and scripts due to ESMValGroup/ESMValCore#999

• Closes Updating recipes/diags to use the new interface for fx variables #2108
• Link to documentation:

---

Before you get started

• ☝ Create an issue to discuss what you are going to do

Checklist

It is the responsibility of the author to make sure the pull request is ready to review. The icons indicate whether the item will be subject to the 🛠 Technical or 🧪 Scientific review.

• 🛠 This pull request has a descriptive title
• 🛠 Code is written according to the code quality guidelines
• 🛠 Documentation is available
• 🛠 Tests run successfully
• 🛠 The list of authors is up to date
• 🛠 Any changed dependencies have been added or removed correctly
• 🛠 All checks below this pull request were successful

New or updated recipe/diagnostic

• 🧪 Recipe runs successfully
• 🧪 Recipe is well documented
• 🧪 Figure(s) and data look as expected from literature
• 🛠 Provenance information has been added

New or updated data reformatting script

• 🛠 Documentation is available
• 🛠 The dataset has been added to the CMOR check recipe
• 🧪 Numbers and units of the data look physically meaningful

---

To help with the number of pull requests:

• 🙏 We kindly ask you to review two other open pull requests in this repository

[valeriupredoi]

thanks lots @sloosvel 🍺 Am surprised this is the only use of fx_variables in a recipe (did grep it myself) 😁

[sloosvel]

I was surprised too! But shouldn't this be merged after ESMValGroup/ESMValCore#999 ? I think the recipe and cmorizers will break otherwise, with the change in the interface.

[valeriupredoi]

indeed - thought that was merged already! Heck, am going straight there now 😁

[valeriupredoi]

also, this would break the recipe anyway in user-mode since the released esmvalcore will be behind until the next release, but I am not a fan of postponing PRs til release of a dep - I know @bouweandela is 😁