Multitemporal analysis for preservation of obsidian sources from Melka Kunture (Ethiopia): integration of fieldwork activities, digital aerial photogrammetry and multispectral stereo-IKONOS II analysis

摘要

An integrated analysis of recent satellite imageries and dated aerial photos demonstrated to be a good investigation tool (Gallo et al., 2009) for the identification of new sites and for the assessment of landscape changes of wide archaeological areas in Ethiopia. In the Melka Kunture archaeological sequence, the obsidian exploitation represents a leitmotiv during the last 1.7 Myr and the first known example of Oldowan utilization (Piperno et al., 2009). The primary nearest obsidian source, the site of Balchit, 7 km North from Melka Kunture, is a domeflow dated to 4.37 ±0.07 Myr (Chernet et al., 1998). Large areas of the Balchit site are covered by secondary obsidian debris resulting from the erosion of primary sources. The proximity between primary and secondary Balchit obsidian sources and the high quality of this raw material, easily available in large quantity, represent a unique condition in the framework of the Oldowan and Acheulian East African sites (Piperno et al., 2009). In order to evaluate the human impact on the multitemporal change of obsidian sources, the land use of the study area, following the CORINE Land Cover Nomenclature, has been classified from aerial photos and IKONOS II imageries, respectively dated to 1972 and 2006. The accurate positioning of the primary and secondary sources and their extent have been measured thanks to the multispectral characteristics and to the high spatial resolution of the available imageries. Satellite scenes, covering an area of about 100 km~2, have been also utilized in stereoscopy for the creation of the new topographic map, at the scale of 1:10,000, and of the Digital Elevation Model (DEM). Images orientation has been performed through the use of Rational Polynomial Coefficients (RPC) which accuracy has been improved by the availability of Ground Control Points (GCPs) properly measured during a DGPS survey. Then, the images have been orthorectified and radiometrically and spectrally enhanced in order to favour the recognition of obsidian presence, in respect with ground observations collected during fieldworks. Photointerpretation and semi-automatic classification processes of images have been performed with the support of spectral signatures of obsidian samples recorded by a FieldSpec Pro spectroradiometer, ranging in the visible-short wave infrared electromagnetic interval (0.4-2.5 mm).