Detection of Building and Infrastructure Instabilities by Automatic Spatiotemporal Analysis of Satellite SAR Interferometry Measurements

Satellite synthetic aperture radar (SAR) interferometry (InSAR) is a powerful technology to monitor slow ground surface movements. However, the extraction and interpretation of information from big sets of InSAR measurements is a complex and demanding task. In this paper, a new method is presented for automatically detecting potential instability risks affecting buildings and infrastructures, by searching for anomalies in the persistent scatterer (PS) deformations, either in the spatial or in the temporal dimensions. In the spatial dimension, in order to reduce the dataset size and improve data reliability, we utilize a hierarchical clustering method to obtain convergence points that are more trustworthy. Then, we detect deformations characterized by large values and spatial inhomogeneity. In the temporal dimension, we use a signal processing method to decompose the input into two main components: regular periodic deformations and piecewise linear deformations. After removing the periodic component, the velocity variation in each identified temporal partition is analyzed to detect anomalous velocity trends and accelerations. The method has been tested on different sites in China, based on InSAR measurements from COSMO-SkyMed data. The results, verified with in-field surveys, confirm the potential of the method for the automatic detection of deformation anomalies that could cause building or infrastructure stability problems.