OSCARS at the RDA 27th Plenary Meeting

RDA 27th Plenary Meeting Banner
Venue
UK Research and Innovation (UKRI), London (UK)
Details
6 October 2026
2026-10-06T00:00:00 - 2026-10-06T23:59:59


OSCARS representatives will contribute to two sessions at the RDA 27th plenary meeting to be held on 6-8 October in London.

Data Quality: Towards Cross-Domain Harmonisation, Provenance-Aware Approaches, Uncertainty Management And Interoperable Practices

OSCARS WP3 co-lead, Romain David, will contribute to the RDA 27th Plenary meeting with the session Data Quality: Towards Cross-Domain Harmonisation, Provenance-Aware Approaches, Uncertainty Management And Interoperable Practices - scheduled on 6 October 2026, from 12:00 to 13:30 BST.

The session aims to:

  1. Present several ongoing initiatives related to Data Quality and the Quality of Data Management.
  2. Identify common challenges and gaps across disciplines and infrastructures.
  3. Explore synergies between EOSC, CODATA, RDA, and other international and ’island our under represented specific’ initiatives (e.g. ISC World Data System, ISO or specific communities).
  4. Discuss and create Concept Notes to the role of provenance, uncertainty, metadata quality, and contextualisation in quality (of data and data management) assessment.
  5. Explore and map cross-domain approaches for harmonising quality concepts, variables, indicators, methodologies and management tools.
  6. Evaluate community interest in establishing a future RDA group dedicated to these topic

Abstract of the session

Data quality criteria are increasingly recognised as foundational components of research ecosystems and essential enablers of FAIR and Open Science practices, interoperability, and the production of reusable and trustworthy data. Approaches to assessing and communicating about data quality remain fragmented across disciplines, infrastructures, and initiatives, due to various sources of heterogeneity (e.g. definitions, methodologies, quality indicators, provenance practices, and governance frameworks)

The emergence of large-scale data aggregation approaches, notably driven by Artificial Intelligence and cross-domain analytics, as well as the integration of heterogeneous data for cross-domain research. This observation / evolution demands robust mechanisms to qualify, contextualise, and communicate the various levels and dimensions of data quality and data management quality across diverse sources and research ecosystems. 

This Birds of a Feather (BoF) aims to bring together communities involved in EOSC and equivalent international initiatives, CODATA, RDA, research and e-infrastructures, technology infrastructures, and domain-specific initiatives to discuss current efforts, identify common challenges, and explore opportunities for convergence around shared concepts, practices, and frameworks for Data Quality and Data Management Quality. The longer-term vision is to survey and characterise approaches to data quality assessment across research disciplines, identify commonalities, and formulate recommendations for communicating data quality through metadata with sufficient richness for both individual researchers and large-scale aggregators to assess the fitness-for-use of datasets.

The session will at least build on:

The BoF will explore whether the community interest and maturity are sufficient to support the establishment of a future RDA Interest Group or Working Group on Data Quality and Quality of Data Management.

Essential Variables: Case Studies In FAIR Commensurability And Interoperability

The other WP3 co-lead, Anca Hienola, representing the ENVRI Science Cluster, will contribute to the session Essential Variables: Case Studies In FAIR Commensurability And Interoperability, scheduled on 6 October, from 14:30 to 16:00 BST.

The session aims to:

  1. List a number of key EV and related activities.
  2. Identify and characterise key components of EV activities, including characterising of EVs, variable description, metadata and semantics, technical and scientific governance.
  3. Articulate ways in which the CDIF4EOSC project can contribute to and support further work to characterise, compare and contract EV initiatives and variable description.
  4. Discuss and articulate a number of possible areas of collaboration, possibly in the context of an RDA IG.

Abstract of the session

The notion of Essential Variables has gained increasing traction in a number of fields of research, particularly those that are vital for Earth observation and with important policy implications. They include Essential Climate Variables (ECV), Essential Ocean Variables (EOV) and Essential Biodiversity Variables (EBV). Meanwhile, in other fields of research, although the term Essential Variables is not always used, there have been comparable initiatives to define the most scientifically significant properties that need to be measured and observed; to describe the appropriate methods and techniques to be used; the expected level of precision, accuracy and spatio-temporal coverage; and to characterise the metadata and semantics for describing the data thus created. Such activities can be observed in crystallography, chemistry, statistics and more. For convenience, we will use the term EV frameworks to describe all these activities, even though there may exist a diversity of scope, motivations and content, which will be important also to examine.

Such EV frameworks, and comparable activities, raise a number of important issues: they are significant in science policy, to focus resources and activity; they are essential to help ensure commensurability of observations, data quality and data Interoperability. There are important corollaries also for scientific and data infrastructures (e.g. what the Global Biodata Coalition calls ‘Core Resources’, those data and information assets, including semantic resources, that are essential to the science). It is also essential to examine the relationship among EV frameworks, the SDGs and other key indicators that are fundamental to many branches of science. 

This BoF will seek to explore a number of perspectives and dimensions of EVs and comparable initiatives:

Science Policy: How are EV frameworks being articulated in a science policy context? How can science policy help advance international coordination around EV frameworks?  

Science: What are the scientific processes for identifying and developing EV frameworks? How can these processes be optimised and supported?

Data, Metadata and Semantics: What are the commonalities across different EV frameworks? What can we learn from EVs and across initiatives about criteria and tests for data quality? How can EVs best be represented semantically and in metadata in different ways? Define minimal metadata requirements to ensure reproducibility and reusability of EVs? These are represented in CIF format using the imgCIF dictionary. What can we learn from EVs for CODATA’s Cross-Domain Interoperability Framework (CDIF) and for approaches such as I-ADOPT or ESIP’s Science on Schema.org (SOSO)?

Although the session will touch on the policy dimensions, the primary focus will be on discussing the role of EVs in relation to data workflows, data, metadata and semantics.

The session will invite a number of activities in relation to EVs, and analogous initiatives, to present their work and to examine to what extent there are points of contact and what the implications may be for data stewardship activities, metadata and semantics. There are links to existing RDA and other activities. CDIF/CDIF4EOSC is placing great emphasis on variable characterisation as a key component of interoperability. In the ocean and environmental domain (ENVRI-Hub NEXT), I-ADOPT is being used to associate observed variables with Essential Climate and Ocean Variables. Inclusion and representation of units is also an important element of semantic representation and interoperability, a topic that is addressed by the CODATA DRUM (Digital Representation of Units of Measure) Task Group.

Examples from the community will include:

  • In X-ray, neutron, and electron diffraction structural analysis, the measured diffracted intensities can be considered as EVs, however, these values are derived from raw diffraction images and are strongly dependent on the experimental setup, described by core metadata parameters essential for reproducibility. The IUCrData journal has established a new section focused on raw datasets and their interpretation and re-usability. 
  • ECV working group in the ENVRI-Hub NEXT project on ECV Use Cases.
  • Proposed standardisation of EOVs specification sheets to support machine readability, interoperability, fast data access and selection
  • I-ADOPT WG
  • Harmonising and identifying essential/common variables across both population and disease longitudinal cohort studies 
  • Climate-health exceedance variables derived from Earth observation datasets (e.g., ERA5, CHIRPS, satellite PM2.5 etc) for interoperable epidemiological analyses and disease surveillance workflows Africa. 
  • SDMX as a meta-standard for EVs and SDGs: how SDMX concepts, codelines, recommendations and ‘structural designs’ relate to EV and SDG frameworks.

The BoF will discuss what the focus of an RDA IG on Essential Variables might be and whether the community interest and maturity are sufficient to support the establishment of such a group. It will feed into a CODATA-convened session on the topic of EVs at the International Science Council’s Midterm Membership Meeting in Beijing on 19 October and the outcomes will be considered by the CDIF4EOSC project.