A physics-based framework for enhancing PSI bridge monitoring
Remote sensing techniques, particularly Interferometric Synthetic Aperture Radar (InSAR), offer a promising, cost-effective solution for civil infrastructure monitoring without the need for on-site instrumentation. Among InSAR methods, Persistent Scatterer Interferometry (PSI) exploits temporally stable scatterers to measure ground displacements with millimetric accuracy. However, the PSI technique faces inherent limitations when applied to complex structural systems such as long-span bridges. In these cases, the deformation may differ from the simple, linear models typically assumed for ground motion, leading to loss of phase coherence and phase ambiguity issues. Consequently, potentially valuable – but low-coherence – scatterers are often discarded from analysis, resulting in an incomplete interpretation of the structure behaviour. This work introduces a novel framework to integrate physics-based structural models (e.g. finite element models) with PSI to overcome these limitations. Rather than focusing on individual pixels, the model accounts for spatial correlation of the persistent scatterers using the structural model of the bridge. The method enables the inclusion of low-coherence Persistent Scatterers that would otherwise be excluded, enhancing the spatial density and reliability of displacement data. The methodology is applied to the Colle Isarco Viaduct (Vipiteno, Italy), a reinforced concrete bridge monitored with multi-temporal COSMO-SkyMed X-band SAR data. The infrastructure is also monitored with topographic survey measurements, which are used in this work as a validation benchmark to assess the accuracy of the results. Results demonstrate that the proposed framework successfully reduces uncertainty in LOS displacement for poorly coherent PSs from approximately 8 mm to 3 mm, within the uncertainty bounds of the benchmark. Furthermore, the enhanced interpretation of low-coherence points provides valuable insights into the bridge structural response and thermal deformation patterns.