Conference Agenda

Geoscientific research yields diverse outputs, including samples, data publications, research code/software and articles. The fragmentation of this knowledge, where different types of research outputs are stored and published in isolated systems, poses a challenge to scientific progress, hindering comprehensive data analysis and collaboration. Additionally, researchers must comply with various data laws, such as the Geological Data Act of 2020, and adhere to Open Science and FAIR principles. The linking of contextually connected research outputs via persistent identifiers (PID) like DOI, ORCID, ROR or IGSN is vital. An interconnected research infrastructure, such as closely collaborating repositories, enhances the value of research outputs and facilitates the overcoming of these challenges.

The Specialised Information Service for Geosciences (FID GEO), funded by the German Research Foundation (DFG), is an example of such an integrated approach. FID GEO facilitates the dissemination of various research results and offers geoscientific communities access to the repositories GFZ Data Services for data and software, and GEO-LEOe-docs for the publication of texts and geological maps.

Moreover, FID GEO provides consulting and training services to assist researchers in effectively managing and linking their research results. These services encompass conference presentations, workshops, and support for the digitisation and online publication of older works.

FID GEO encourages its community to publish and describe all their research outputs. By integrating data and text repositories, FID GEO streamlines research workflows, enhances the dissemination of scientific contributions, and supports compliance with necessary principles.

The National Research Data Infrastructure for Earth System Science (NFDI4Earth) aims to create a networked infrastructure connecting different research data sources (repositories, data bases) to overcome the challenges associated with Research Data Management (RDM) in Germany. Researchers collaborate in international research teams, resulting in interconnection between national and international activities. Thus, a core element of the initiative is the ambition to harmonize the research data landscape on a national level and to connect them with international initiatives. The key areas of interest are to increase the interoperability between different research data domains, metadata standardization, controlled vocabularies, application programming and the setup of different service interfaces. NFDI4Earth would like to support researchers in various aspects such as discovering and exploring relevant data sources, publishing and curating data and addressing research data management challenges. NFDI4Earth focuses on being attractive for the user by having a user-friendly entry point (OneStop4All), services such as a living handbook, user support network, educational resources, and an academy for early career scientists, the knowledge hub as the technical backend, as well as participative opportunities. NFDI4Earth is open to promote common metadata standards, to support, motivate and enable the wider ESS community appropriately to move towards FAIR (Findable, Accessible, Interoperable, Reusable) and Open Science Data principles. NFDI4Earth initiatives are in line with the European Open Science Cloud (EOSC), an ecosystem of research data and related services; such as multidisciplinary projects FAIRCore4EOSC and FAIR-EASE. Thus, NFDI4Earth advances the cultural shift towards FAIR and open RDM.

The ongoing digitization enables the development of new methods for data-driven research and of large research infrastructures (RI) across the Earth and environmental sciences. The increasing demands for RIs to enable seamless data integration and visualisation requires the harmonisation and interoperability of data formats, and the use of agreed metadata standards. Especially for data intensive disciplines in geophysics and geodesy, disciplinary metadata and data standards are important and already in place and widely adopted which makes their integration in new RIs easier than for small and highly variable datasets from the long-tail communities.

In addition, it becomes increasingly relevant to make data discoverable in the internet (via their metadata) and to digitally connect research outputs (articles, data, software, samples) with each other and with the originating researchers and institutions – in unique and machine-readable way. The use of persistent identifier (like DOI, ORCID, ROR, IGSN) and descriptive linked data vocabularies/ontologies in the metadata associated with research outcomes are strongly supporting these tasks.

Research data repositories, especially domain repositories, are experts for this Domain repositories are digital archives that manage and preserve curated research data from specific scientific disciplines. The metadata associated with the DOI-referenced objects is specific for their domain and richer than generic metadata supposed to describe data across many scientific disciplines. Their metadata for data discovery is provided in machine-readable formats (XML, JSON) following international standards (e.g. DataCite, ISO 19115/INSPIRE) and include all information for the development of knowledge graphs. As such they are much more partners than opponents of RIs.

Data is only useful if you can get your hands on it. When you have mountains of data to wade through, you need the best, most efficient methods of finding precisely what you are looking for. The easiest way to look for a needle in a haystack is to use a magnet. A good search strategy is that kind of magnet, helping you find that needle of desired data in the haystack of countless sets!

Classical metadata schemes might not be sufficient to guide one to all relevant data sets. The name of the data creator or the year of publication might not be known. Many times data sets are linked to research papers but this does not simplify the search because research papers are just another haystack. A combination of keywords or title words are rarely unique and may yield numerous hits.

Using classical metadata and viewing long query result lists is time-consuming, but even with this method there is a risk of not spotting relevant data. Either one does not know the best keywords or the creator did not label the data set with the best keywords. This presentation makes suggestions on how to find geoscientific data better and gives examples of web based information systems.