Remote sensing techniques for monitoring cultural heritage sites

Climate change is likely to have a direct impact on tangible cultural heritage. Cultural heritage sites are already experiencing the impact of variations in temperature, precipitation, atmospheric moisture, and wind strength, along with rising sea levels and shifts in the frequency of extreme events. Leveraging remote sensing tools presents an opportunity for the effective surveillance and detection of potential threats to Cultural Heritage sites, along with monitoring the material deterioration in monuments. Since monuments are not isolated in the geographical space, assessment and evaluation of changes in their broader area are important because they serve as warning signals to the concerned stakeholders and facilitate them to take measures for preventing CH asset damages. Satellite data after appropriate processing provide significant “background” information by pointing out a) hazards with a slow or gradual onset in the broad area of the CH monuments and facilitating change assessment including ground deformation and land cover changes and b) assessing damage on both the surrounding region and the monuments after events like floods, landslides, earthquakes, etc.

Within HYPERION project (https://www.hyperion-project.eu/), RS-based methods have been developed for routine monitoring of the CH sites and were tested at four pilot sites. Routine monitoring includes displacement and land cover change detection maps of the broad area for all the pilot sites that are studied within the Hyperion project (city of Rhodes, Granada, Venice and Tønsberg), flood monitoring maps, three-dimensional models for all the CH assets, and deterioration and material loss estimation for specific parts of the facades of two monuments in the pilot site of the city of Rhodes, the Fort of Saint Nikolas and the Roman bridge in the Rhodes island.

To this end, a) advanced methodologies using PS and SBAS functionalities and Convolutional Neural Network architectures have been applied on satellite data aiming to produce reliable land deformation and land cover change detection maps, respectively, b) time series analysis and classification have been employed to identify changes in backscattering and to map flood occurrences, c) point clouds, light models, texture models and sections have been created in order to obtain a detailed 3D representation of the assets, and c) hyperspectral processing methods have been employed for fast and efficient assessment of the material deterioration level.

The effectiveness of the developed methods has been evaluated through their implementation on the pilot sites and disseminated to the scientific community through relevant publications. This paper presents a comprehensive overview of the results they have achieved and highlights the capabilities of remote sensing as a valuable tool in preserving Cultural Heritage.

Acknowledgement:
This work was implemented in the framework of the HYPERION project (H2020-LC-CLA-2018-2 H2020 program under GA 821054). Funding for participation in the conference has been provided by the HORIZON-CL2-2022-HERITAGE-01 program, with grant agreement number 101095253.