Contents

How to…

How to…

Get access to the collection

Anyone working/studying in the Australian climate science community with an NCI account can access the collection by requesting access to project ia39 here.

Find data

The collection can be found on NCI systems at /g/data/ia39/aus-ref-clim-data-nci/. You can search the subdictories manually to find what you need or make use of our intake-esm catalogue.

A complete dataset listing is also available in the sidebar of this book and there’s a repository corresponding to each dataset (where you can find information and ask questions) in our GitHub Organisation.

Request a new dataset

Associates of the Australian Climate Service (ACS) or of partner organisations can request that a dataset be added to the collection by creating a new issue here.

Assess a new request

The size, management complexity and number of potential users of a dataset will be considered when assessing a new request. If the number of potential users is high and the size and management complexity is low, a dataset can typically be added without a formal committee meeting. If the dataset is complex to manage, requires a lot of storage and/or is only of use to a small number of people in the community, the committee will meet to consider the request. A requestor may be asked for additional information in support of their request, and in some cases may be asked to help with the ongoing maintenance of the dataset.

Please note that this project is not obliged to satisfy any data request. We’ll do our best to meet the needs of the community and to assess requests in an open and transparent manner, but will sometimes be limited due to storage, staff time or ACS project constraints.

Add a new dataset

For each new dataset create:

  • a folder with the dataset acronym in lowercase letters in /g/data/ia39/aus-ref-clim-data-nci/, with a data and a code subfolders.

  • a new repository in the aus-ref-clim-data-nci github organisation.

  • the repository should be then cloned in the code subfolder.

Permissions

The dataset directory should be owned by the data custodian. The data folder is owned by ia39 for accounting reasons, and acls are used to give ia39_w writing access:

setfacl -R -m g:ia39_w:rwx code/

The code folder is owned by ia39_w with rwx permissions for the group, other can read but not write or execute. For help with posix and acsl permissions see ACDG Governance guide.

Each repository will contain wherever possible and relevant:

  • download code file/s with sufficient comments to make sure the scope and use of the code is clear

  • a markdown README.md file with the following subsections:

    • overview: Summary of the dataset

    • data download: How (and where from) was the data downloaded?

    • data location: Where was the data downloaded to?

    • license: Details of the license associated with the dataset.

    • data citation: Is there a DOI for the dataset?

    • references: Are there any academic papers describing the dataset?

    • acknowledgement: Any acknowledgement statement the dataset authors request that people include in their papers.

    • author note: Any other notes from the dataset authors.

    • assistance: A link to the issues page of the repo so where people can ask for assistance.

  • license for the code as a separate file

  • log file updates_log.txt with details of first download and subsequent updates to the files where suitable

If using an existing publically available code there is no need to replicate it but clear references to the original source (i.e. github repository) and how it is used should still be detailed in the README file.

Update a dataset

The frequency of dataset updates depends on the specific dataset. We will try wherever possible to follow the original dataset schedule. Where regular updates are necessary we will use the accesdev Jenkins server to manage scheduling. In this way any user will be able to check when a dataset was updated last, if the update was successful and when the next update is scheduled. Collection managers will have permissions to add, edit and run tasks. The accessdev Jenkins uses the NCI LDAP access protocol so they can use their NCI credentials to login. To just view the scheduled tasks is not necessary to login. Updates will be also documented and shared in the update_log.txt file available in each repository and locally in the <dataset>/code folder.

Retire a dataset

We are still in the process of developing a retirement policy. This will eventually be based on the recommendations of the Australian Climate Data Guide (ACGD) Climate Dataset Guidelines working group.

previous

Governance

next

Currently available