Multispectral remote sensing for post-dictive analysis of archaeological remains. A case study from Bronte (Sicily)

In recent years, the use of remote sensing in archeology has rapidly grown, having being proven particularly useful for landscape archaeology investigations. The quality of results obtained through remote sensing mainly depends on the number of bands that can be exploited, the spatial and spectral resolution of imagery and the methodology used. In this work, adopting a post-dictive approach, multispectral remote sensing data from World-View 2 (WV-2) were used in order to detect archaeological features in the volcanic area of Bronte (Sicily, Italy), a town 40 km from Catania. The selected area lies in the western slope of Etna, inside the boundary of Etna Park. Here, the archaeological research is affected by specific problems. The geology and the spontaneous vegetation of this fascinating environment represents one of the biggest issues. The entire research was conducted using free and open source software (FOSS) within a low-cost logic that allows study of landscapes using limited budget. The strategy developed in our study is directed towards a post-dictive approach to archeological data and remote sensing techniques. We analyzed an area where archeological buildings are now visible, but WV-2 multispectral images have been deliberately selected prior to the archaeological excavations carried out in 2015. In this way, we can evaluate and test instruments and techniques at our disposal, emphasizing weaknesses and strengths. The final results illustrate the benefits and challenges of performing multispectral imagines for archaeological investigations. The use of some of the most promising vegetation indices, especially Normalized Difference Vegetation Index (NDVI) and Simple Ratio Index (SRI), is analysed and discussed, while the post-dictive approach suggests that the more useful bands are undoubtedly NIR_1 and RED_EDGE.