Multi-temporal UAV-based multispectral analysis for vegetation-related risk assessment in archaeological areas

Abstract. This study investigates the potential of UAV-based multispectral imagery to support vegetation management in archaeological contexts where structures are preserved underground and direct intervention is limited. The analysis focuses on the grassland covering the buried Roman theatre within the Archaeological Park of Castelleone di Suasa (Italy), exploring the relationship between vegetation spectral responses and the Deteriogenic Index (DI), an indicator that indirectly expresses vegetation-related pressure on buried remains. Vegetation indices derived from multispectral data acquired at two distinct moments within the same vegetative season were analysed in relation to plot-based DI values obtained from floristic surveys. Results show that vegetation dynamics, rather than single-date conditions, provide more consistent information, with inter-seasonal indices highlighting vegetation persistence associated with higher deteriogenic pressure. The calibrated relationship was used to derive a spatial proxy of DI, offering a synthetic and non-invasive tool to support spatially informed and adaptive vegetation management strategies in archaeological parks and contributing to the development of integrated information systems for the analysis and sharing of multidisciplinary environmental data.