Anomaly identification in historical modular structures using satellite data: Methodology and application to Portico di San Luca, Bologna

This paper presents a novel methodology for detecting and characterizing structural anomalies in historical modular structures using satellite-derived displacement data. The proposed framework monitors short-term correlations among statistical features of different structural segments and identifies anomalies as individual deviations from the collective behavior. Segments exhibiting such deviations are classified as anomalous and further characterized with the aid of a simplified parametric model that requires only basic geometric information. This strategy enables the transfer of knowledge from simulated damage scenarios to real-world cases without the need for detailed mechanical modeling. The methodology is applied to the Portico di San Luca in Bologna, Italy, a historic arcade recently designated as a UNESCO World Heritage Site. Detected anomalies are validated against average velocity maps, a conventional product in satellite-based structural health monitoring. Whereas velocity maps require multi-year datasets to estimate long-term trends, the proposed method computes an anomaly index at each acquisition, enabling near-real-time detection while explicitly filtering out environmental effects acting uniformly across the structure. The results demonstrate the potential of satellite-informed structural health monitoring at the regional scale, offering a scalable, non-invasive, and cost-effective tool for the preservation and preventive maintenance of cultural heritage assets. • Statistical features are extracted from interferometric displacement data. • Anomalies are detected if the behavior of a structural segment deviates from the others. • Environmental drivers are accounted using autoregressive models. • Simplified numerical models enable anomaly characterization via transfer learning. • Settlement was identified in a segment of the Portico di San Luca, Bologna.