A NEW ROLE FOR PHYSICAL SAMPLES IN THE GEOSCIENCES BIG DATA WORLD

To date, most efforts in the curation and management of physical samples have been focussed on supporting the long tail community and enabling linkage to the physical samples that were used to generate analytical data such as geochemical data (trace elements, isotopes, geochronology, etc.), spectral measurements and physical rock property measurements (e.g., magnetic susceptibility, density, porosity, permeability, etc). More recent efforts on the digitisation of physical samples are working towards enabling online linkages to be made between the digital representation of the samples, the analytical data and the research publications derived from these samples.

However, much geoscientific data today is derived from high volume remotely sensed data sets (e.g., airborne data, autonomous instruments, satellite data). These remotely sensed data collections are proxies of real world features and need to be calibrated and validated. To mount field campaigns to go and collect the required samples would be prohibitively expensive. With growing accessibility of digital representation of physical samples and their associated data, through efforts such as IGSN and iSamples, these samples can be linked to existing measurements on literally millions of physical samples from long tail collections collected around the world over the last 200 years. Increasingly, researchers are turning to these historic collections to enable them to now be reused and repurposed in support of the calibration and enhancement of modern high volume Geoscience Big Data sets. In some cases, old historic samples are being reanalysed using newer high precision analytical instruments to futher enhance their relevance and value.