Sampling and specimens: potential application of a general model in geoscience sample registration

Sampling is a key element of observational science. Specimens are a particular class of sample, in which material is retrieved from its original location and used for ex-situ observations and analysis. Specimens retrieved from difficult locations (e.g. deep ocean sampling, extra-terrestrial sampling) or of rare phenomena, have special scientific value. Material from these may be distributed to multiple laboratories for observation. For maximum utility, reports from the different studies must be recognized and compared. This has been a challenge as the original specimens are often not clearly identified or existing ids are not reported. To mitigate this, the International Geologic Specimen Number (IGSN) provides universal, project-neutral identifiers for geoscience specimens, and SESAR a system for registering those identifiers. Standard descriptive information required for specimen registration was proposed during a SESAR meeting held in February 2011. The standard ISO 19156 'Observations and Measurements' (O&M) includes an information model for basic description of specimens. The specimen model was designed to accommodate a variety of scenarios in chemistry, geochemistry, field geology, and life-sciences, and is believed to be applicable to a wide variety of application domains. O&M is implemented in XML (as a GML Schema) for OGC services and we have recently developed a complementary semantic-web compatible RDF/OWL representation. The GML form is used in several services deployed through AuScope, and for water quality information in WIRADA. The model has underpinned the redevelopment of a large geochemistry database in CSIRO. Capturing the preparation chain is the particular challenge in (geo-) chemistry, so the flexible and scalable model provided by the specimen model in O&M has been critical to its success in this context. This standard model for specimen metadata appears to satisfy all SESAR requirements, so might serve as the basic schema in the SESAR repository. However, more and more geospatial data discovery systems are being built on the ISO 19115 metadata model. The design of 19115 was most strongly influenced by the requirements of maps and imagery (series), but may be treated as a generic metadata standard and used for a variety of scientific artefacts by assigning an appropriate 'ScopeCode'. Most of the requirements for SESAR specimen description can be accommodated in 19115, potentially allowing standard metadata tools like GeoNetwork and the NOAA metadata system to provide an interface to SESAR specimens. What is significance of these alternatives? The O&M Specimen model provides a specific view of specimen metadata, suitable for data entry or presentation to scientists, while the 19115 model provides compatibility with generic metadata systems that include descriptions of resources of many types. Understanding the relationships and transformations between these representations is a critical step towards efficient sharing of data and documentation across multiple communities.