Attribute-Enrichment of Satellite InSAR Deformation Measurements Based on Contextual Information

Abstract:

Over the past decades, Interferometric Synthetic-Aperture Radar (InSAR) has proven itself useful for monitoring surface displacements. Nonetheless, due to the complexity of InSAR observations, expert knowledge is still required for their interpretation and applicability. This research is set to build a system that enables the end users to collect and interpret fully processed InSAR data sets based on their user requirements. To this end, InSAR data sets from different missions (e.g. Sentinel-1, Radarsat-2, TerraSAR-X) are enriched with contextual information. In a spatial database, each point entry denotes its coherent reflection and is provided with its ’inherent’ attributes, i.e. its coordinates, elevation, displacement time series, and processing-specific meta data. Contextual information is assigned to the points with the intent to either determine their origin in a 3D space or to specify any (set of) predetermined condition(s), such as their presence in a risk area, in an area of particular interest, or in any other type of classified environment. As a basic functionality, the attribute-enriched InSAR database serves as a platform for fast and efficient querying. Moreover, spatio-temporal patterns in the deformation of objects, or even entire areas, can be identified through exploratory data analysis or via other search algorithms. The system can be applied in any case-specific context as well as in a broader sense, e.g. to connect spatial data sets from different satellite missions and acquisition techniques. The dynamic workflow described in this contribution will demonstrate its use through a study of surface deformation in the Netherlands, i.e. for sinkhole detection in the former mining areas of Limburg and for large-scale subsidence of gas depleted areas of Groningen.