RITARS-11-H-UVA Sinkhole Detection , Landslide and Bridge Monitoring for Transportation Infrastructure by Automated Analysis of Interferometric Synthetic Aperture Radar Imagery Final Report

During the two year project, a series of radar satellite data were collected and processed by TRE Canada scenes using interferometric synthetic aperture radar (InSAR) technology into ground displacement measurements. These displacement data, with accuracy below 1 cm and approaching 1 mm, provided the initial data with which to study phenomena affecting transportation in Virginia. The phenomena analyzed included subsidence due to sinkhole formation, movements due to landslides and rockslides, and bridge settlement. The study area was comprised by a 40km by 40km region near Middlebrook, Virginia. Additional data outside of Virginia (from Vancouver, Canada and Wink, Texas) provided by TRE Canada proved useful in prototyping and validating the image analysis algorithms. A focus of the project was the development of image analysis algorithms that take the InSAR data as input and provide outputs of detections that can be used in a decision support system (DSS) to identify potential hazards to transportation. Two main theoretical approaches were explored: a graph-theoretic approach and parametric approach. In the graph theory approach, regions of subsidence were identified by an optimization process. This approach however was limited and did not allow for an easy integration of the main feature offered by the InSAR acquisition: the displacement time history for each scatterer. The second more generalized parametric approach exploited the temporal dimension as well as the spatial data. This approach is based on the availability of models describing both the spatial and temporal behavior of the geophysical features of interest. The model parameters are used to generate a multidimensional space that is then scanned with user-defined resolution. For each point in the parameter space, a spatiotemporal template is reconstructed from the original model. This template is then used to scan the point cloud data set for regions matching the spatiotemporal behavior. This new parametric approach provides the flexibility necessary to allow extensibility to other geophysical phenomena of interest. Photogrammetry, LiDAR, as well as traditional surveying methods were used as comparison to the InSAR-driven results and these ground studies confirmed and validated the results achieved from remote sensing. This final report details a number of case studies and inspections performed by the Virginia Department of Transportation (VDOT) including cases of sinkhole formation, bridge settlement and rockslides. In terms of automated geohazard detection (as provided by the newly developed algorithms operating on the InSAR data acquired over Virginia), the ground studies show that about 78% of the …