BUILDING A CUSTOM RELATIONAL DATABASE FOR THE CURATION OF SEDIMENTARY GEOCHEMICAL DATA IN THE LAB

The Neoproterozoic-Paleozoic interval was a time of considerable change in terms of marine redox conditions, and the subject of intensive research in recent decades. However, analyses of long-term changes are limited, in part, by insufficient data and a lack of data curation. Here we discuss how we are bringing together new and existing geochemical datasets (in particular, major and trace element concentrations, iron speciation, and metal isotope values) in a local relational database, based in part on the data models of the British Geological Survey, EarthChem, and collection management systems (e.g. Specify, Arctos). In addition to geochemical data, we record details of analytical methods, geographic location, geological age, lithology, sedimentary structures, degree of weathering and metamorphism, depositional environment, geological region and associated body and trace fossils. The structure allows for flexible querying across large datasets, quality control, and the establishment of good metadata standards in the lab. By using similar controlled vocabularies, storing reference ids (e.g. Paleobiology Database, Macrostrat and British Geological Survey ids), and including IGSNs (international geo sample number), our design allows for connections to be made with complementary datasets. Our aim is to facilitate robust statistical analysis of redox conditions through time and enable comparisons with new and existing compilations of animal body size, diversity, and ecology, while ultimately generating an accessible, well-curated discipline-specific geochemical dataset of lasting value to the Earth history community.