Issues of Sampling and Representativeness in Large‐Scale LiDAR‐Derived Archaeological Surveys in Mediterranean Contexts

Abstract

Landscape‐scale LiDAR‐based studies are becoming increasingly prevalent in archaeology, mainly focusing on detecting archaeological sites to create datasets for spatial analysis. However, the representativeness of these datasets in accurately reflecting the surviving distributions of archaeological sites has often been overlooked. This paper discusses issues of sampling and representativeness in LiDAR‐derived datasets, particularly within the scope of large‐scale landscape studies in Mediterranean contexts. Drawing insights from the Ancient Hillforts Survey, which analysed 15 296 km2 in south‐central Italy, the study examines the variability in the visibility of different site typologies in open‐source but low‐resolution LiDAR data. Through an examination of hillforts, platform farms, settlements, field systems, traces of Roman centuriation, and transhumance routes, the paper highlights significant variability in the identification and mapping within and across different site types. Recognizing the need to account for this variability in the development of spatial analysis, the paper discusses the use of sampling areas to address this variability. This approach aims to effectively mitigate potential biases in analysis, emphasizing the necessity for nuanced methodologies in interpreting LiDAR data for archaeological research.