Within the framework of disaster risk management, this article proposes an interdisciplinary method for the analysis of multiple natural hazards, including climate change’s influences, in the context of cultural heritage. A taxonomy of natural hazards applicable to cultural heritage was developed based on the existing theoretical and conceptual frameworks. Sudden-onset hazards, such as earthquakes and floods, and slow-onset hazards, such as wetting–drying cycles and biological contamination, were incorporated into the hazard assessment procedure. Future alteration of conditions due to climate change, such as change in heat waves’ duration, was also taken into account. The proposed hazard assessment framework was applied to the case of the Historic Centre of Rethymno, a city on the northern coast of the island of Crete in Greece, to identify, analyze, and prioritize the hazards that have the potential to cause damage to the center’s historic structures. The assessment procedure includes climate model projections, GIS spatial modeling and mapping, and finally a hazard analysis matrix to enable the sharing of a better understanding of multiple hazards with the stakeholders. The results can facilitate decision making by providing the vulnerability and risk analysis with the nature and spatial distribution of the significant hazards within the study area and its setting.

Hyperspectral images can highlight crop marks in vegetated areas, which may indicate the presence of underground buried structures, by exploiting the spectral information conveyed in reflected solar radiation. In recent years, different vegetation indices and several other image features have been used, with varying success, to improve the interpretation of remotely sensed images for archaeological research. However, it is difficult to assess the derived maps quantitatively and select the most meaningful one for a given task, in particular for a non-specialist in image processing. This paper estimates for the first time objectively the suitability of maps derived from spectral features for the detection of buried archaeological structures in vegetated areas based on information theory. This is achieved by computing the statistical dependence between the extracted features and a digital map indicating the presence of buried structures using information theoretical notions. Based on the obtained scores on known targets, the features can be ranked and the most suitable can be chosen to aid in the discovery of previously undetected crop marks in the area under similar conditions. Three case studies are reported: the Roman buried remains of Carnuntum (Austria), the underground structures of Selinunte in the South of Italy, and the buried street relics of Pherai (Velestino) in central Greece.

Ground penetrating radar (GPR) is a well-established technique used in archaeological prospection and it requires a number of specialized routines for signal and image processing to enhance the data acquired and lead towards a better interpretation of them. Computer-aided techniques have advanced the interpretation of GPR data, dealing with a wide range of operations aiming towards locating, imaging, and diagnosis/interpretation. This article will discuss the novel and recent applications of machine learning (ML) and deep learning (DL) techniques, under the artificial intelligence umbrella, for processing GPR measurements within archaeological contexts, and their potential, limitations, and possible future prospects.

This article focuses on the possible drawbacks and pitfalls in the GPR data interpretation process commonly followed by most GPR practitioners in archaeological prospection. Standard processing techniques aim to remove some noise, enhance reflections of the subsurface. Next, one has to calculate the instantaneous envelope and produce C-scans which are 2D amplitude maps showing high reflectivity surfaces. These amplitude maps are mainly used for data interpretation and provide a good insight into the subsurface but cannot fully describe it. The main limitations are discussed while studies aiming to overcome them are reviewed. These studies involve integrated interpretation approaches using both B-scans and C-scans, attribute analysis, fusion approaches, and recent attempts to automatically interpret C-scans using Deep Learning (DL) algorithms. To contribute to the automatic interpretation of GPR data using DL, an application of Convolutional Neural Networks (CNNs) to classify GPR data is also presented and discussed.