Presentation #109.05 in the session Open Science and Data Tools (Oral Presentation)
Introduction: ESA’s Planetary Science Archive (PSA) is the home for all data from the Agency’s planetary missions. It is a single archive providing cross-mission search and data access to holdings in both PDS3 and PDS4 formats. This paper reports on recent and planned updates to PSA services.
The newly-released PSA version 7 includes a new user interface offering a more modern and responsive design, as well as new features and performance improvements. In addition, the APIs offered are being consolidated and updated to offer access to all versions of the data, and a new way to analyse and visualise the data (ESA DataLabs) is being rolled out.
PSA 7: The new version of the PSA user interface (see Fig. 1) provides some long-requested features, including the ability to share a URL to a complete search (restore UI state) and to a single product. For PDS4 data, search by common PSA attributes is possible (e.g., observation ID and type). Advanced search has been aligned between the TAP API and user interface; both now use ADQL and can access the same database fields.
APIs: Traditionally the PSA has supported the PDAP protocol, and more recently, added support for a subset of the EPN-TAP data model. In preparing for the release of the new user interface, the entire PSA database has now been exposed via the TAP API, not just those key meta-data used by EPN-TAP. In this way power users can run more advanced queries. Even more powerful search is enabled by the PDS registry and API, which allows the ingestion and search of PDS4 products by any meta-data. PSA is currently populating the registry with our PDS4 data.
Coming soon: Features currently in development, for release this year, include:
- A download manager (shopping basket) which uses the power of PDS4 links to serve the user relevant content (e.g. documents, calibration data, source products)
- Map views for Phobos, and preparation for a map view for Mercury.
DataLabs: As well as the various interfaces available to access ESA’s space science archives, the DataLabs project will bring a new way of interacting with the data. It allows users to build and run pipelines and interactive applications which run close to data. Early use cases are focusing on data tutorials in Jupyter Notebooks (Fig 2), but prototypes are also being built for data visualization and, later, more ambitious projects using large quantities of data for machine learning and related applications. The current state of DataLabs and its integration with the PSA will be shown.