-
Clarifying the DRS: do a final check to ensure that the CV in use is aligned with other models and as best with CORDEX as possible (noting that the CORDEX specifications don't exist yet for CMIP6).
-
Metadata: does your data contain the required metadata? (see ACS Data Standards)
-
Licensing: does your data contain a license in the metadata? This should be CC-BY v4.0 unless otherwise agreed or directed by ACS leadership.
-
Documentation and README: there will need to be a README file along with your release data, that preferably points to documentation of the data, and links to any published papers.
-
Assigning a data custodian: who will be responsible for the ongoing management of this data, including requests, updates, etc?
-
File system management & permissions: ACLs to set up correctly (globally readable, but with an ia39_w mask for writing). To confirm:
- Set owner to data custodian:
chown -R $data_custodian_id $directory - Set group to ia39:
chgrp -R ia39 $directory - Set permissions to user writable, group & other readable:
chmod -R 755 $directory - Set ACLs to enable writability by ia39 writers group:
setfacl -R -m g:ia39_w:rwx $directory
- Set owner to data custodian:
-
Advertising: will you be advertising the dataset widely/locally?
-
Intake catalogue: are your data catalogued in an intake-esm catalogue for ease of access and use?
-
Post-processing code under version control: is the code and schemas used to generate the final datasets in the ACS Github, or otherwised version controlled?