Release v2.2.0 things to include after the Feb. 1st, 2021 freeze

[valeriupredoi]

Things to include on a "heck yeah, we really need this" basis:

General faff

• should do (even if DKRZ is missing at the mo) Add instructions how to use esmvaltool from HPC central installations #841
• should do (nice user-friendly) Update install.rst #848
  - possibly Refactored mask regridding for irregular grids (fixes #772) #865
• must do Dataset fixes for: ACCESS-ESM1-5, CanESM5, CanESM5 for carbon cycle #947 (IPCC relevant)
• must do Add option to exclude input cubes in output of multimodel statistics #978
• must do Set remove_preproc_dir to false in default config-user #979
• must do Restore task summary in debug log #981
• possibly Change the order of the publication steps #984 if merged/wanted in this release by Javi
• must do (release dox) Add checkbox in PR template to assign labels #985
• must do (bug!) Bugfix: Fix units for MSWEP data #986
• must do Move fiona to be installed from conda forge #987 - fiona needs some caressing
• must do (re-add fiona to setup.py) re-added fiona in setup.py #990

Dataset Fixes

• Fixes for KIOST-ESM (CMIP6) #904
• if Javi says "Si, segnor" Fixes for AWI-ESM-1-1-LR (CMIP6, piControl) #911

[valeriupredoi]

@jvegasbsc I am torn if we would want to include more dataset fixes or not - some are ready, some are ready(ish) too - what do you think? I'd say do them in for the next release (apart from the ones that are really needed - like the IPCC one above)

[jvegreg]

Data fixes are quite safe: they should not be a problem as they are usually very focused. PRobably we are safe to add anything that does not include an AllVars

[valeriupredoi]

@jvegasbsc successfully released, this goes to bin 👍