Use of Satellite Remote Sensing for Detecting Archaeological Features: An Example from Ancient Corinth, Greece

Abstract: During the last few decades, satellite remote sensing has proven to be an important non-invasive method for archaeological research in order to detect ancient sites and manage existing ones. Archaeologists have tried in the recent past to embed Geographic Information Systems (GIS) and image processing techniques in their research as non-destructive approaches, which can allow a wider perception of archaeological landscapes and predict past behaviors. Adding to traditional means of visually identifying archaeological features from satellite imagery, multi-spectral satellite images, by offering bands that cover a wide range in the electromagnetic spectrum, can also help archaeologists to identify potential sub-surface features and map exposed archaeological remains. Through the application of vegetation indices, archaeologists are using differentiations in the spectral signatures of vegetation that could indicate potential man-made structures beneath soil while classification techniques have been employed mostly for the study of above ground features. However, their implementation usually takes place in a rural setting while there are often difficulties in validating the results. In this thesis, the site of Ancient Corinth, in Greece and its existing archaeological record gives us the opportunity to apply the aforementioned, established image processing techniques in a semi-rural environment and compare them with known excavated features. This can help us evaluate the application of these techniques for detecting and mapping buried and exposed archaeological remains and potentially add more information to the archaeological background of the region.