[DATA REQUEST] 20km regional projections for CORDEX-CMIP6 from the Queensland Future Climate Science Program

[rbeucher]

Description of the data product

This dataset includes projections at 20km, formatted to meet the CORDEX-CMIP6 data standards. The 20km projections were derived from the 10km projections.

Location of the data product on Gadi

Catalog available in /g/data/ig45

Checklist

Add a "x" between the brackets to all that apply

• This data product is stable (unlikely to change substantially or move)
• This data product is of use to the broader community
• This data product is documented: link
• This data product is licensed under Creative Commons Attribution 4.0 International
• Those who want to access this data can be added to the project that houses it

[marc-white]

Have requested membership of this project to investigate.

[marc-white]

Please confirm this YAML is suitable for the dataset (hte pre-built catalog does not list variables, frequencies, etc.):

name: cmip6_ig45
experiment_uuid: c7021d1e-7ba2-11ef-beb5-000007d3fe80
description: 20km regional projections for CORDEX-CMIP6 from the Queensland Future Climate Science Program
long_description: >-
  This dataset includes projections at 20km, formatted to meet the CORDEX-CMIP6 data standards. 
  The 20km projections were derived from the 10km projections.
model:
- CMIP6
frequency:
- 
variable:
- 
nominal_resolution:
- 
version: 
contact: NCI
email: [email protected]
reference:
license:
url: https://geonetwork.nci.org.au/geonetwork/srv/eng/catalog.search#/metadata/f7465_8388_5100_7022
parent_experiment:
related_experiments:
-
notes: 
keywords:
- cmip

[rbeucher]

I had a look at the esm catalog... It has not been generated properly.
I'll raise an issue with NCI

[rbeucher]

Contacted NCI today 02/10/2024

[rbeucher]

The catalog has been fixed.

[charles-turner-1]

I've just noticed when building the metadata.yaml file that this dataset contains the 10km projections in addition to the 20km projections (see attached image).

I'm not sure if this is intentional or the 10km projections are just being brought along for the ride, so to speak? I've included reference to them in metadata.yaml but kept the focus on the 20km dataset

[marc-white]

I guess it doesn't really hurt to have them both? Given that the 20km projections are derived from the 10km, I'm assuming there's probably someone who wants to do comparisons/sanity checks between the two.

[charles-turner-1]

Sounds sensible. I'm ready to drop the PR in on this one but will wait until later today - I've layered these changes on the type annotations I added in #212 - can cherry pick out the translators if it seems sensible.

[marc-white]

@charles-turner-1 I think cherry-pick out the type annotations - that'll get them into main faster, as well as reduce #212 to being only directly relevant changes.

[marc-white]

The underlying data for this request appears to have been updated in mid-late November, due to data errors being discovered. The updated data set is no longer compatible with the CordexTranslator, and is causing issues with attempts to build a catalog using v1.0.0b2 (#230 ).

[charles-turner-1]

This will also require adding an xfails marker to the CordexTranslator test in our end to end tests until fixed - working on that now