Towards a Collaborative Online Workspace and Unified Standards for Geochemical Data

Most major geochemical datasets are derived by aggregation of individual sets of analyses on samples collected from spatially located point sources. As the variety and capacity of geochemical instruments continues to increase, the lack of agreed international standards mean that geochemical data are increasingly heterogeneous and harder to aggregate. There have been numerous attempts to combine disparate geochemical analyses into seamless national and international data sets but most geochemical data bases only store the finalised results as presented in tables in publications. Scientifically robust aggregation will rely increasingly on easily accessible provenance metadata on how the raw data was collected and processed. We propose an automated capture system that follows the fundamental scientific methodology. It starts with the instrument that captures the data, uses web services to make standardised data reduction programs more widely accessible, and finally uses internationally agreed data transfer standards to make geochemical data seamlessly accessible online from a series of internationally distributed certified repositories. As science is built on the principle of reproducibility, independent researchers should be able to collect the same data and analyse it to produce similar results. Current practices in geochemistry do not make this feasible. Often there is no standardisation of analytical instruments for the same technique and some instrument manufacturers use proprietary data formats. Further heterogeneity is introduced when the raw data are processed as different methods for calibration and quantification of the data are developed (some of which can be proprietary third party software packages). The Australian National Data Service (http://www.ands.org.au/) is funding a range of data capture solutions to ensure that the data creation and data capture phases of research are fully integrated to enable effective ingestion into research data and metadata stores at the institution or elsewhere. They are developing a national discovery service that enables access to data in institutional stores with rich context. No data is stored in this system, only metadata with pointers back to the original data. This enables researchers to keep their own data but also enables access to many repositories at once. Such a system will require standardisation at all phases of the process of analytical geochemistry. The geochemistry community needs to work together to develop standards for attributes as the data are collected from the instrument, to develop more standardised processing of the raw data and to agree on what is required for publishing. An online-collaborative workspace such as this would be ideal for geochemical data and the provision of standardised, open source software would greatly enhance the persistence of individual geochemistry data collections and facilitate reuse and repurposing. This conforms to the guidelines from Geoinformatics for Geochemistry (http://www.geoinfogeochem.org/) which requires metadata on how the samples were analysed.