Modelling geoarchaeological resources in temperate alluvial environments: The capability of higher resolution satellite remote sensing techniques

Abstract

Temperate river systems cover large regions of the earth's surface and provide challenging environments for the detection of archaeological remains. The deposition of alluvium above, and interbedded with, archaeological deposits means conventional forms of archaeological prospection such as geophysical survey or aerial photography are often ineffective. Consequently, archaeologists have turned to the construction of deposit models to evaluate the archaeological potential of various landforms. However, despite the high capacity of satellite multispectral data to identify alluvial landforms and archaeological deposits, the contribution that it can make towards deposit models has only received limited attention. This research assesses the capability of higher resolution satellite multispectral imagery to identify alluvial landform assemblages and subsequently facilitate the modelling of geoarchaeological resources. As image enhancements are often required to improve interpretation or classification, it includes a quantitative evaluation of 45 different processes, comprising two spectral separability measures on paired regions of interest (ROIs) from the Lower Lugg Valley, Herefordshire, UK. The results define optimal image processing techniques for the detailed geoarchaeological interpretation of alluvial environments and demonstrate that relatively simple visualisation methods (e.g. composite images) and summative techniques (Principal Components Analysis) offer some of the most effective approaches.