University of Colorado
Metadata for the 5th Coupled Model Intercomparison Project (CMIP5) was collected by a large online questionnaire that used a community-developed metadata schema for climate models, the Common Information Model (CIM). The questionnaire was written in python, relied on realm-specific (e.g. atmosphere, ocean) controlled vocabularies, and was mostly hardwired. Since CMIP5, various organizations have come forward requesting that questionnaires be generated for their projects and unique datasets. Modifying the CMIP5 questionnaire for these new groups is time consuming and clumsy.
A better solution is enabling projects to generate project-customizable questionnaires. The CIM Questionnaire is written in Django (a Python web framework) and is being used to support numerous projects including the 2012 Dynamical Core Model Intercomparison Project (DCMIP), as well as the 2013 Quantitative Analysis of Downscaling Workshop.
Figure 1: Screenshot of a CIM Questionnaire for a climate model showing the complex component hierarchy and tabbed structure for various questions.