Innovative method for provenance studies in cultural heritage: A new algorithm based on observables from high-resolution Raman spectra of red ochre.

Abstract

Red ochre, typically derived from iron oxides and hematite, has been used since Pleistocene times for a range of different applications, practical as well as symbolic, including cave paintings and use in prehistoric burials. The importance to discover new methods for provenance determination, based on non-destructive portable techniques, represents a new challenge in the field of diagnostics of cultural heritage. This study presents the data obtained from the analysis of several non-flaked tools and ochre-stained bones, showing evidence of ochre processing at the Mesolithic site of S'omu e S'Orku in Sardinia (Italy). To investigate the provenance of the ochre (hematite phase) found on a massive stone from the site and also used to cover the bones, we propose three distinct approaches derived from high-resolution Raman spectra of ochres, aiming to identify the maximum number of observables that can be reconducted to unicity criteria. The reliability of this method enables the development of an automatic algorithm of Artificial Intelligence able to recognize the provenance of raw materials used in a range of activities. Furthermore, this study sheds light on one of the earliest and most distinctive Mesolithic burials uncovered in Sardinia to date, providing valuable insights into the human colonization of the island and the symbolic practices of its inhabitants during the Holocene epoch.