Community Standards for Modeling Science

Establishing community standards to improve computational modeling in a changing world.

The OMF coordinates and administers the development or adoption of pan-community standards that accelerate knowledge scaffolding among modeling scientists, and promote the creation and use of more reusable, replicable, interoperable, and reliable models. These standards include:

  • Standards for model documentation sufficiently comprehensive as to enable others besides a model developer to understand and potentially replicate the purpose and operation of a model.
  • Standards to facilitate the reuse of a model by others beyond the context in which the model was originally developed; for computational models this includes reusability regardless of the operating system or language in which a model was originally developed. 
  • Standards to enable potential interoperability among different models; this includes a standards for a pan-community external API and vocabulary (ontology).

The OMF coordinates and administers the certification of models that meet these standards and the assignment of badges or other public indicators of such certification to the general modeling community. Models are certified as meeting the following standards:

  1. Discoverability and Accessibility
  2. Documentation
  3. Reusability and Replicability
  4. Interoperability 

The definitions of those standards are a living document, and suggestions by the broader modeling community to update them will be considered by the OMF standards committee, and become part of the standard once approved. 

The OMF also encourages and coordinates peer review of models by the scientific community and other evaluation activities, like model intercomparison programs, that enhance the reliability and value of models. 

In addition to standards applicable to models, the OMF also develops, adopts, and administers community-wide standards for ethical practice by modeling scientists. In order to encourage knowledge exchanges and scaffolding, all modeling scientists should publicly credit (normally through citation) the work of any other scientists that they use. In order to maintain the reliability and public reputation of modeling science, practitioners must document accurately model purposes, operations, inputs, boundaries, assumptions, and results. Modeling scientists are also strongly urged to make all models transparent and open to external examination, and to make their models openly accessible to others. 

These standards for models and modeling science practice are disseminated to all member organizations and publicly displayed on the OMF website.