2021-10-27, 10:45–11:00, Grand Ballroom
An effort is on-going to make astronomy data FAIR: Findable, Accessible, Interoperable and Reusable. The Reusability principle is particularly subtle as the question of trust is raised. Provenance information on the data is essential to ensure this trust, and it should come with the proper granularity and level of details. The effort to specify the provenance of astronomy data is shared with current and future large observatories in the context of the ESCAPE European project and the International Virtual Observatory Alliance (IVOA). We organized discussions on common approaches across all wavebands. In particular, Cherenkov astronomy provides complex data based on the detection of Cherenkov light generated in the atmosphere by air showers induced by energetic cosmic particles. Unlike current experiments such as H.E.S.S., MAGIC or VERITAS, the next generation Cherenkov Telescope Array (CTA) will be an open observatory with public access to its high level science data products. The complex treatment of raw Cherenkov data implies to attach a detailed description of the high level data and their provenance.
We developed a prototype platform to connect the first H.E.S.S. public test data release to the Virtual Observatory through an IVOA ObsTAP service: relevant metadata is mapped to the IVOA ObsCore data model main fields and additional specific metadata is provided (F, A). The high level data follows a proposed standard data format for gamma-ray astronomy (GADF) to ensure wider interoperability (I). We finally designed a provenance management system in connection with the development of pipelines and analysis tools for CTA (ctapipe and gammapy), in order to collect rich and detailed provenance information, as recommended by the FAIR principles (R). The prototype platform thus implements the main functionalities of a science gateway, including data search and access, online processing, and traceability of the various actions performed by a user.
FAIR standards for astronomical data