Conference Agenda

First studies showcasing the potential of satellite-derived digital elevation models (DEMs) to search for archaeological tells in Near and Middle Eastern archaeological landscapes date back to the early 2000s. Since then, free and open access global DEM datasets at medium resolution such as NASA’s 90 m Shuttle Radar Topography Mission (SRTM) surface model have been increasingly exploited by archaeologists to map tells on a supra-regional scale, and thus analyse past settlement patterns. However, in the specialist literature there is little to no evidence that landscape archaeologists have investigated the improvements brought by higher resolution satellite-derived DEMs, as they were made available by space agencies. To understand how these datasets may support archaeological surveying, we test two high-resolution DEMs generated with (1) interferometric synthetic aperture radar (InSAR) and (2) stereo-photogrammetry (i.e. the two methods typically used for DEM generation), and assess their performance in comparison with openly distributed datasets (i.e. 30 m SRTM DEM and the Advanced Land Observing Satellite World 3D–30 m - AW3D30). We selected the 10 m posting InSAR-derived DEM generated from 3 m resolution StripMap HIMAGE mode images acquired by the Italian Space Agency’s COSMO-SkyMed SAR constellation, and the 5 m posting stereoscopic Cartosat-1 Euro-Maps 3D DEM made available through ESA’s Earthnet Third Party Missions programme and ad-hoc call for R&D applications. The demonstration was run at regional scale in the Governorate of Wasit in central Iraq, where the literature suggested a high density of sites, despite knowledge gaps about their location and spatial distribution. The enhanced observation capability of COSMO-SkyMed DEM was found advantageous to detect both well preserved and levelled or disturbed tells, standing out for more than 4 m from the surrounding landscape. The mapped tells were then compared and cross-validated with those detected using the Cartosat-1 dataset. Combined exploitation of the two DEMs allows improving the knowledge of type, distribution and condition of local archaeological deposits, also in the context of contemporary land use changes and threats for conservation. Archaeological heritage in Wasit is currently at risk of vanishing due to natural erosion and weathering, encroachment of anthropogenic activities (e.g., ploughing, infrastructure projects, modern settlement and dam construction) and looting. DEM integration with Google Earth time-lapses (where available at suitable resolution), CORONA KH-4B tiles, 1950s Soviet maps and Copernicus Sentinel-2 multispectral imagery, enabled the identification of looting incidents and tells affected by anthropogenic disturbance (e.g., road and canal constructions or ploughing). While the results of our experiments contribute to the current vivid research on Iraqi archaeological heritage and its challenges for conservation, the developed methodology may stimulate further exploitation in archaeological landscapes with similar characteristics elsewhere, and the future development of semi-automated site and looting detection approaches.

Since the adoption of the World Heritage Convention (1972), modern technologies have significantly changed the way our society behaves and operates, with an increased demand for energy, fast and reliable communications, etc. Some current technologies (excesive contruction, deforestation, etc.) might contribute to negative impacts on heritage sites, for example through climate change and/or excessive tourism; however, modern digital technologies can also be extremely beneficial for heritage activities. In this paper, we focus on how modern digital geo-science and geo-technology can support heritage authorities’ daily work. We introduce herein the concept of Digital Geoheritage. By this terminology we refer to the data, methods and technologies used to collect, distribute, store, analyze, process, and present georeferenced data in support of heritage applications. Our main research aim is to keep the complexity of “Digital Geoheritage” among the experts in geomatics and remote sensing, deriving easy to understand results which can then help heritage authorities to discover and understand the enormous that all these technologies can provide as support in their heritage activities.

This research case, implemented through an interdisciplinary scientific approach, originally aimed to support the preservation, restoration and management of a cultural heritage site; however, it was later expanded to also support archaeological research, stability risk assessment, planning, design, education, dissemination and promotion. The use of digital geo-sciences for the benefit of the local Maya communities living around a heritage site is also illustrated.

As far as data for our research activities we have been using some satellites from the Copernicus family, high-resolution satellite data, UAVs with standard digital camera and/or infrared camera. For some image data processes, we have used Artificial Intelligence to develop software to facilitate the process.

The presentation will illustrate how Earth Observation and Remote Sensing in general is being used to satisfy the requirements of the Mexican National Authorities who are partners and co-authors of this paper.

Archaeologists commonly use airborne LIDAR technology to produce 3D models of structures, even when obscured by a forest canopy. However, this technology has a high cost, both from the plane itself and from the processing of the LIDAR point cloud. Furthermore, this technique can only be used over limited regions. This paper proposes a technique that uses SAR satellite imagery to identify man-made structures hidden by a forest canopy. To do so, we exploit the Ascending and Descending passes of Sentinel-1 so that we obtain two images of the candidate site but from different sight directions. Because of cardinal effects, a large enough building will sign differently from the comparatively isotropic forest canopy it is obscured by. Practically, the technique is based on the ratio of backscattered intensity from these two illumination angles and is well adapted for large areas. The advantages and shortcomings are discussed for the specific case of Sentinel-1 SAR images over two Maya archaeological sites in Central America. Our analysis shows that SAR satellite imagery might provide a free, global-scale way of preselecting sites with large or tall structures to complement LIDAR technology.

This study examined the application of LiDAR data from the Alianza REDD+ initiative, initially collected for forest monitoring, to visualize and semi-automatically map archaeological structures in the Yucatan region of Mexico. A key advantage of LiDAR is its ability to penetrate vegetation, revealing structures not visible in aerial photographs or satellite images. Our research involved processing the LiDAR data into Digital Surface Models (DSMs) and applying various visualization techniques, such as the Enhanced v3 Multi-scale Topographic Position (e3MSTP), to improve the detection of archaeological features lying within forest cover. Following the visualization process, deep learning computer vision techniques, specifically the Mask R-CNN model, were utilized to automate the mapping of these features using tools in ArcGIS Pro for data preparation. The model was trained on ArcGIS Online with annotated datasets of terrain visualizations from areas with ancient Maya settlements, focusing on detecting and segmenting structures like buildings and ring-shaped features such as ovens and cisterns. The results indicated that the Mask R-CNN model effectively detected and segmented buildings, achieving a balance between precision and recall. However, for ring-shaped structures, the model showed high precision but lower recall, suggesting cautious predictions with some missed detections. This highlights the need for further optimization, particularly in increasing recall for ring-shaped structures. In conclusion, the research demonstrated the potential of re-analyzing LiDAR data using terrain visualization tools and deep learning models for archaeological studies, which could help landscape archaeologists to gain valuable insights into ancient civilizations. Future research should focus on expanding the dataset, developing more objective methods for training and test data collection, and exploring advanced training methods to improve the detection and segmentation of archaeological structures.