Read the Inteview with the Astronomy & Particle Physics Thamatic Pilot to discover the latest updates on OSTrails pilot studies. Explore their progress in integrating open science principles and advancing research assessment. This month we had the pleasure of speaking with Baptiste Cecconi, astronomer at Observatoire de Paris.

- Baptiste Cecconi

"With the Astronomy pilot in OSTrails, we will enhance the community open science practices, with more efficient data management, common interfaces to our registries and fine-tuned FAIR assessments tailored to our ecosystem."

-Can you briefly introduce your organisation? How do they contribute to EOSC?

Observatoire de Paris is the largest astronomy institution in France, gathering a third of the French astronomy community. Its teams participate in many national, European, and international projects (ground-based observatories, space-borne missions, numerical modeling, etc.). Observatoire de Paris is also a leading institution in astronomy interoperability alliances: at the time of writing 4 of the 10 working leads of the International Virtual Observatory Alliance are from the observatory, and the same level of engagement is also true with the International Planetary Data Alliance and the International Heliophysics Data Environment Alliance. Observatoire de Paris is also a member of the EOSC Association and has been testing EOSC provided tools since 2018. Thanks to those activities and the participation in the ESCAPE project, we are now in two EOSC projects (FAIR-IMPACT and OSTrails), as well as two OSCARS cascading grant projects. For us, EOSC is an opportunity to enhance our open science practices, as well as to train our community in scalable and open digital infrastructures.

-What are you most excited about in OSTrails? What are you looking forward to?

Astronomy is often considered a mature community thanks to its long-lasting open science practices and interoperability policies. Despite over two decades of development of protocols and tools that have provided the astronomy community with a very efficient open data infrastructure, we remain in a siloed situation, addressing the challenges of open science in very specific ways. For instance, FAIR assessment tools are currently not able to evaluate the FAIRness of our resources. In OSTrails, our main goal is to enhance our open science practices, for example, by integrating our resource registry as a Scientific Knowledge Graph (SKG) with other OSTrails tools and by developing FAIR metrics and profiles, ensuring that our specific tools are included in the FAIR assessment.

-How is planning, tracking and assessing research being realised in your scientific domain?

Major astronomy projects involve partners from many countries from all over the world. This is mainly driven by the cost of these infrastructures, which can’t be afforded by a single country. On the one hand, international organisations such as the ESO (European Southern Observatory), or the ESA (European Space Agency) provide a framework for planning research, which can be a burden in terms of documentation for project members. On the other hand, tracking and assessing is not really organised on a large scale. There are citation and acknowledgment policies for publications, and funders are crawling repositories to find statements mentioning the projects. The assessment of the resources is done by the IVOA registry group, monitoring the interfaces’ conformance to the specifications. This applies only to resources served using IVOA protocols. Furthermore, linking scholarly publications to datasets and services providing access to those datasets is not yet a common practice in our community, although provenance metadata is gaining interest in many projects. 

-Can you provide some details on your pilot's main actors, services and priorities? How will your pilot adopt the results of OSTrails?

The Astronomy pilot in OSTrails is developed at Observatoire de Paris. It covers the three pillars of OSTrails: plan, track and assess. It is also splited into two sub-pilots (radio astronomy on the one hand, and high-energy astrophysics on the other hand), so that we make sure that the prototyping is generic enough. 

The radio astronomy sub-pilot will be developed using the MASER repository, whose research products cover the three astronomy alliances’ scopes. The high-energy astrophysics sub-pilot concerns the CTA (Cherenkov Telescope Array) data management, which has very specific needs in terms of provenance management.

The details of the pilot are described below according to each OSTrails pillar.

Plan. At Observatoire de Paris, we started to develop a Data Management Plan (DMP) template back in 2016, with the EPN2020RI project. From the beginning, we were willing to make our DMP useful, so that it is not just a report to funders and thus seen as a burden to the science teams. We included many aspects and metadata that are required for us to efficiently manage the life cycle of the research data, e.g., we not only list the specific protocol used to share the data, but we also include metadata, which allows us to deploy the service. 

Track. Astronomy data alliances have their own registries of resources, with specific interfaces to access them. Celestial astronomy has its own specific database and protocol and is sharing metadata concerning data access services. Planetary astronomy mostly shares archived products, focusing on the technical reusability of data files in the long term, rather than interoperability. Heliophysics provides an XML-based registry listing various types of products (from instruments to numerical data), but without a specific API. OSTrails gives us the opportunity to set up a common interface and abstract those independent registries as knowledge graphs. 

Assess. As previously mentioned, we have tools to assess the conformance of our interfaces to our specifications. It is part of the IVOA governance to require validators for any proposed recommendation. This ensures that our infrastructure is up and running. In OSTrails, we have two objectives in FAIR assessment: the first is to foster the improvement of our resources’ FAIRness, using the commonly accepted FAIR criteria (e.g., use of persistent identifiers, use of ontology/thesauri URIs, etc); and the second is to include in the FAIR assessment our specific interfaces and policies. 

-Ongoing activities and Next Steps? 

We are now testing DSW and have started implementing knowledge models and templates matching our requirements. Our first findings and results are very promising, and other teams at the observatory are already interested in implementing the tool. 

Concerning the connection to our community registries, we have not yet started actual developments, but we are involved in the RDA SKG-IF working group. We have proposed to add a new type of resource, which is a “Data Access Service”, since that’s mostly all we have in the IVOA registry. 

I was nominated as a FAIRsharing champion for Astronomy last month, and I am now reviewing and adding records in FAIRsharing that we will use for our FAIR assessment tests. The connection to our service validation tools is yet to be developed.

We realize that our pilot in OSTrails is ambitious, but we really think that most of the puzzle pieces are already there. Most of the work will consist in assembling the right pieces and using the right glue to achieve our goals and providing a useful set of tools for our community.