Control Vocabulary Software Designed For CMIP6

The Coupled Model Intercomparison Project Phase 6 (CMIP6) coordinates a number of intercomparison activities and includes many more experiments than its predecessor, CMIP5. In order to organize and facilitate use of the complex collection of expected CMIP6 model output, a standard set of descriptive information has been defined, which must be stored along with the data. This standard information enables automated machine interpretation of the contents of all model output files. The standard metadata is stored in compliance with the Climate and Forecast (CF) standard, which ensures that it can be interpreted and visualized by many standard software packages. Additional attributes (not standardized by CF) are required by CMIP6 to enhance identification of models and experiments, and to provide additional information critical for interpreting the model results. To ensure that CMIP6 data complies with the standards, a python program called “PrePARE” (Pre-Publication Attribute Reviewer for the ESGF) has been developed to check the model output prior to its publication and release for analysis. If, for example, a required attribute is missing or incorrect (e.g., not included in the reference CMIP6 controlled vocabularies), then PrePare will prevent publication. In some circumstances, missing attributes can be created or incorrect attributes can be replaced automatically by PrePARE, and the program will warn users about the changes that have been made. PrePARE provides a final check on model output assuring adherence to a baseline conformity across the output from all CMIP6 models which will facilitate analysis by climate scientists. PrePARE is flexible and can be easily modified for use by similar projects that have a well-defined set of metadata and controlled vocabularies.