Skip to content

Research Practices

The CDR Forecasting Stack project is committed to responsible, transparent, and reproducible research. The following practices guide research conduct, authorship, publication, and software development across the project.

Authorship and Contributions

  • Authorship will follow the Vancouver Convention.
  • Author contributions will be described using the CRediT taxonomy.
  • Authorship and contribution decisions should be discussed openly and revisited as projects evolve.

Transparency, Reproducibility, and Open Research

  • Transparency and reproducibility are core requirements of the research and publication process.
  • Whenever possible, publications should be made openly available, either through open-access journals or institutional or disciplinary repositories (e.g., arXiv or Earth ArXiv).
  • Data supporting publications should be made openly available whenever feasible. When full data release is not possible, reduced or derived datasets sufficient to support published results should be shared.

Software, Data, and Licensing

  • Code developed specifically within the CDR Forecasting Stack project should be hosted under the CDR Forecasting Stack GitHub organization.
  • Software developed in collaboration with, but not exclusively for, the project should be open source and accessible without authentication or fee barriers.
  • Private repositories may be used during active development.
  • Licensing decisions remain with the original authors but should be consistent with host institution policies and adhere to the FAIR principles or the Open Source Definition.
  • Recommended licenses include MIT for software and Creative Commons CC-BY for non-code products such as text, figures, and websites.