Among biological disciplines, e.g., the Global Biodiversity Information Facility (GBIF, https://www.gbif.org) provides the most comprehensive and prominent search portal for species occurrences including collection object and observation data. As GBIF focusses on biodiversity only, the presentation of fossils is inadequate and data describing abiotic specimens like rocks and minerals are completely missing. In order to overcome this, the Geoscientific Collection Access Service (GeoCASe, http://www.geocase.eu) a domain specific portal aggregating data using the ABCDEFG standard, was developed. Here, geoscientific objects can be published and researched in a convenient way.

The maintenance of the portal, adding new features as well as further developments for adapting to the rapid changes in web technologies are ongoing challenges. In order to increase the acceptance of the portal and provide a sustainable service, it is necessary to focus future efforts primarily on demands arising from the scientific community itself.

For collecting user demands, specialists from the geosciences were asked to participate in an online survey about geoscientific collection data, data publication, and cross-institutional data search (see Suppl. material 1 for the survey template and Suppl. material 2 for all given answers). If not specified, we always give the frequency as the absolute amount of answers given to a particular question.

Twenty-five scientists, of which 48% were curators, completed the questionnaire. They are associated with 22 institutions located in seven different countries. The institutions hold a variety of geoscientific collection objects with minerals and rocks being the most common specimens (Fig. 1).

Figure 1.  

Types of collection objects. Given are the frequencies of particular objects the survey participants’ institution or department holds. Other objects include soil, crystal models, historical instruments, boreholes, etc.

Around 50% of the participants are making collection object metadata available. The data are accessible via international data portals such as GeoCASe, GBIF, and Europeana (https://www.europeana.eu/portal); national data portals like the German museum-digital (https://www.museum-digital.de), or their own institutional websites. The usage of data standards varies from 'no standard', to 'agreements within the department or institutions', to using international data standards like Darwin Core or ABCDEFG. However, one third of the participants are not aware of existing conventions for the publication of object metadata (answered 'none' or left question blank).

The perceivedvalue of data portals for cross-institutional search varies among potential audiences. The participants name primarily the scientific community and secondarily the institution as the groups most benefiting from existing data portals. For daily individual work the importance of portals differs among the participants between high and low. Based on the survey, published geoscientific object data is of minor importance for the general public. See Fig. 2 for further details.

Figure 2.  

Estimated value of cross-institutional search possibilities through data portals on geoscientific object (meta)data for different target groups. Shown is the frequency of answers for different scopes and their respective importance (see legend).

Thereasons for publishing or not publising data on geoscientific collection objects through a data portal were also asked about. According to the responses received, higher accessibility and higher visibility are the main reasons for publishing through a data portal (Fig. 3a). Other reasons like 'special policies', 'nice to have', or 'because others do' are only of minor importance. On the other hand, concerns regarding security, bad data quality and potential misapplication of the data make researchers and curators refrain from publishing particular data sets (Fig. 3b).

Figure 3.  

Reasons for publication of (meta)data on geoscientific collection objects (a) and not publishing particular datasets through a data portal (b). Shown is the frequency of answers per reason. Note that the y-axes are scaled differently.

Another question focused on the minimum functionality that a data portal needs to cover and which optional features it could provide in order to increase its value for daily scientific work. Here, multiple differences between each geoscientific sub discipline emerge and several participants emphasize that the minimum information needed depends on the type of collection and scientific task performed.

For paleontological objects, the following properties were mentioned: scientific name / taxonomy, common name, type status, details of the specimen (which part of the skeleton), locality (i.e. country), age, geochronology, lithostratigraphy, quantity, etc.

For rocks and minerals, participants listed common name, synonyms or historical names, chemical composition, locality (including details such as mine, administrative unit, or country), geological horizon/age/locality, color, year entered to the collection (historical purposes), where the rock was used (architectural requests), accession (collector, year of collection), including thesauri such as of mindat (https://www.mindat.org), etc.

For meteorites, the name is unique and all information on a specific meteorite is linked to its name. However, the unique reference number, parent-daughter / sibling relationships, mass and storage in the collection are considered as required fields for a data portal.

In general, the inventory number, the housing institution, and a contact person responsible were identified being essential metadata. If available, a photograph of the object would also increase the value. The most desirable minimum information is the name and collection locality as this will enable most searches carried out by external researchers and the public. The functionality of a respective portal should be similar to that of GBIF. Further demands are a basic search function for all fields (e.g. thematic and regional searches), fuzzy searches, queries for classification, descriptive statistics, and an integrated loan service*2.

The last part of the questionnaire focused on the awareness and usage of the Geoscientific Collection Access Service (GeoCASe). GeoCASe is known by around fifty percent of all participants (Fig. 4a), whereas only four institutions were already providing data to the portal. Most participants would like to use the service but require more information on GeoCASe(Fig. 4b). The majority of participants is willing to support the improvement of the portal (Fig. 4c) and would like be part of the GeoCASe community (Fig. 4d).

Figure 4.  

Awareness and usage of Geoscientific Collection Access Service (GeoCASe). Shown is the frequency of answers regarding the awareness of GeoCASe (a), the data provision through the GeoCASe portal (b), the readiness to support the progress of improvements (c), and the interest of being part of the GeoCASe community (d).

The Museum für Naturkunde, Leibniz Institute for Evolution and Biodiversity Science hosted an international workshop focusing on data standards in geosciences on May 29^th and 30^th 2017. During the workshop geoscientists, curators of geoscience collections, and data base developers shared their experiences using data exchange formats in geosciences. The workshop participants were selected based on their scientific affiliation to paleontology (first day) or petrology, mineralogy and meteorites (second day). Specialists with overlapping expertise were welcome to participate on both days.

Both days were structured similarly. Starting with a short personal introduction of all participants, presentations focusing on recent developments and the digital status of important geoscientific collections from institutes worldwide followed (see https://abcd.biowikifarm.net/wiki/Events:WorkshopEFG2017 and Suppl. material 3). The subsequent session delved into the data standard ABCDEFG with a presentation describing the workflow of publishing an institutional collection in the GeoCASe portal using EFG in particular, and highlighted current developments of this common biodiversity data standard. In the last part of the workshop, the additional needs and requests relating to a data exchange format based on use cases was collected from all participants in an open discussion. Furthermore, fundamental elements being vital for the publication of collection data related to different geoscientific disciplines in general (paleontology, mineralogy, petrology, and meteorites) were identified.

Following the thematic talks on the first day, input of all workshop participants regarding the development of a data portal and common data standard in geosciences was collected in an open discussion session.

The workshop revealed not only differences between geoscientific disciplines, but also between collections and institutions with respect to publication of collection object associated information. Although data standards were known by most participants, the majority of them neither publish data using common exchange formats nor do they publish in domain specific portals. Some institutes, however, make images of their collection objects and / or a subset of associated metadata available on their website.

Researchers from all disciplines shared the opinion that rocks and fossils are strongly related with each other and collections objects composed of both should always be described together. Therefore, a data standard should provide the option to refer to related elements e.g. the host rock of a fossil or the mineral structures associated with a fossil. This is of course also true for objects composed of several rock types or minerals.

In contrast to other geoscience collection objects and natural history objects in general, meteorites have a unique name and all information (data, locality, observation etc.) is linked to this particular name. All data is freely available in the reference database of the Meteoritical Society (https://www.lpi.usra.edu/meteor/metbull.php). An institutional inventory number, weight, size, local collection storage, and associated multimedia objects, are however important properties for institutional and research purposes as well as the broader public and thus should be available in any exchange format or data standard used.

In addition, the importance of publically available information also differs among user groups. Scientists ask for different facts compared to the broader public or the creative industry. The color of minerals for example, is only of minor importance for geoscientists but of high importance for an artist planning an exhibition. Given the heterogeneous group of participants of the workshop, the requirements of various potential user groups of geoscience object related data could be collected. Scientists, collection managers, and data base developers exchanged their experience. In addition to the demand for publically available data, requests of past users of their institutional collections were also taken into account. The elements relevant for the publication of paleontological object metadata (Table 1) and the collection of important metadata of rocks, minerals, and meteorites (see above) will be incorporated in the application schemata for the different disciplines.

During the workshop, the participants were asked to identify object metadata which need to be publically available. A couple of the terms identified for one subject are of general importance for all natural history collection objects and will of course be part of all application schemata.