Infrastructure Health Monitoring Using SAR Tomography

Static structures, such as bridges and dams, are in need of continuous structural health monitoring due to the risk of collapse or destabilization, which leads to casualties and material loss. From economical point of view, the frequent inspection of these structures is cost-ineffective and time consuming. This shortage is overcome by using remote sensing systems such as synthetic aperture radar (SAR) systems, that do not require the in situ positioning of sensors and observe the Earth on a very large spatial scale and at regular time intervals. In particular, Tomographic Synthetic Aperture Radar (TomoSAR) has proved to be an effective tool for structure health monitoring. It is a multi-dimensional imaging technique that exploits stacks of interferometric SAR images acquired with slightly different view angles and at different times. TomoSAR is able to analyze the temporal deformations and the millimeter displacements on buildings. The small movements that occur due to thermal sensitivity of the structure can be also captured, so that the coefficient of thermal expansion, which is a fundamental parameter in health structure monitoring, can be computed. The widespread deterioration and some recent collapses of infrastructures have highlighted the importance of developing early warning monitoring strategies, devoted to identify structural problems before they become critical. A problem to be considered is that effective monitoring techniques have to rely on approaches capable of identifying few representative points, which describe well the dynamical behavior of the structure. A TomoSAR technique recently developed, based on a multi-step binary hypothesis GLRT test, has shown a good capability of selecting reliable control points on the structure for estimating its deformations. In order to demonstrate the potentialities of this technique for structure health monitoring, two tall structures will be considered. The results obtained using different SAR sensors, such as COSMO-SkyMed, Sentinel-1A and TerraSAR-X, are presented.