Near Infrared imaging spectroscopy for raw materials characterization: the example of a Mesolithic dwelling site in Northern Sweden

MOBIMA is a research project carried out at Umeå University which aims to develop non-destructive and image-based field techniques for chemical analysis. Within MOBIMA we test applications of Near Infrared imaging spectroscopy in archaeological research. A dataset of lithics from a dwelling site dated to mid-late Mesolithic at Lillsjön, Ångermanland, Sweden was analyzed using a sisuCHEMA pushbroom shortwave infrared hyperspectral imaging system (spectral Imaging Ltd, Oulu, Finland) for acquiring images from 1000 to 2498 nm at intervals of 6-7 nm. The spectral data were centered prior to principal component analysis and the Savitzky-Golay algorithm was applied for noise reduction. We could classify raw materials according to their spectral signature and ascribe objects to different types of quartz, quartzite, slate and flint. This preliminary study shows how Hyperspectral NIR (Near Infrared) imaging could be used in provenance studies as fast non-destructive screening method for characterizing raw materials.