Mine dans le Morvan
L’étude “Establishing archaeological prospectivity maps of ancient mines on the basis of EDA and fractal model applied to geochemical data” par F. Monna, C. Gourault, E. Camizuli, R. Nedjai, G. Hamm, F. Cattin, B. Bohard, C. Petit, J.-P. Guillaumet, P. Alibert, présentée à l’IGC 2012, Brisbane. [ppt].
Projet “Identification et impact des sites miniers abandonnés sur les écosystèmes aquatiques et terrestres actuels”.
Résumé: The exploitation of the polymetallic ore deposits in the Morvan Massif (France) is now attested for the Iron Age by archaeological excavations. Several recent environmental studies have even suggested that mineral resources may have attracted populations as early as the Bronze Age. However, in a so vast (5000 km2) and woody region, it is very difficult to undertake systematic pedestrian prospection. This study aims to delineate geochemical anomalies from the background in order to target more specifically the areas where mineral substances are abundant, and hence to drastically restrain the size of the area to be prospected. First, a pilot site of 30 x 30 km2 has been selected, where 2543 stream sediment samples have been previously collected and analyzed by the BRGM for their chemical composition in the 1970s-1980s. Using the “rgr” and “StatDA” packages for the free R software, geochemical data are treated by Exploratory Data Analysis and by more recent fractal techniques (Concentration – Area plots) in order to produce prospectivity maps, introducing or not tectonic constrains. The efficiency of the models is then checked using the 119 past and modern mining sites already discovered by pedestrian prospection of the field study. Prospectivity maps also reveal the nature of the metal exploited for mining areas where field evidence is lacking. Once validated for the pilot study, the best model is applied to the whole Morvan Massif whose the database exceeds 25,000 analyzed samples (http://sigminesfrance.brgm.fr). It provides efficient guidelines for future archaeological prospection.