Detection and identification of archaeological features using aerial LIDAR data in a forested environment (Châtillon-sur-Seine, Côte-d’Or, France).

International audienceIn the field of archaeology, a range of processing techniques has been developed this last decade for the visualization and analysis of LIDAR elevation data. Traditional processing methods are based on illumination techniques (e.g. hill-shading, Sky View Factor, topographic openness…) or on morphometric analysis (e.g. slope, profile curvature, aspect, local relief model…). All these techniques have shown their potential for identifying archeological features, especially in forested environment where photo interpretation is not usable. In this work, such techniques have been applied to detect archeological features from the processing of LIDAR data that were acquired in 2012 in the forest of Châtillon-sur-Seine (France). The study area is covered by sets of protohistoric to medieval dry-stone structures which were investigated during 10 years of GPS prospection. Each feature was classified according to a morphometric typology . Three main classes of morphology were defined: linear forms, terraces and punctual features. Among the different processing techniques used to analyze LIDAR data, the calculus of local slope map appears to be one of the most effective techniques to identify the dry-stone structures. Local variations of slope values facilitate the detection of new structures and their assignments to one of the typology defined and validated by the GPS prospection. However, the calculus of local slope map does not permit to define if the local relief of these features is in positive or negative elevation. To overcome this gap and in addition to the local slope index, we have calculated the topographic positive openness index. This index expresses the degree of “dominance” of a location on an irregular surface, and permits negative elevation feature to be recognized