Northeastern Section - 49th Annual Meeting (23–25 March)

Paper No. 2
Presentation Time: 8:25 AM


VACCARI, Andrea1, BRUCKNO, Brian S.2, HOPPE, Edward3, ACTON, Scott1 and CAMPBELL, Elizabeth4, (1)University of Virginia, Department of Electrical and Computer Engineering, P.O. Box 400743, Charlottesville, VA 22904, (2)Virginia Department of Transportation, Materials, 811 Commerce Rd, Staunton, VA 24401, (3)Virginia Center for Transportation Innovation & Research, 530 Edgemont Road, Charlottesville, VA 22903, (4)Virginia Department of Transportation, 1401 E. Broad St, Richmond, VA 23219,

As part of the USDOT-funded research program RITA-RS-11-H-UVA, “Sinkhole Detection and Bridge/Landslide Monitoring for Transportation Infrastructure by Automated Analysis of Interferometric Synthetic Aperture Radar [InSAR] Images,” the authors completed the a pilot study in which they developed a computational approach aimed at the early detection and evaluation of potential geohazards within a point cloud dataset obtained from processed InSAR data. The technique was applied to the detection of sinkholes within an active 40x40 km data frame located in the Valley and Ridge Province in Virginia.

The analysis, based on the detection of a specific spatio-temporal model describing incipient sinkhole behavior, was used to scan a 10 million point dataset for regions where the spatio-temporal behavior matched the model, providing as output a geo-referenced map indicating the quality of match. This map was then converted to a risk map where fastest growing features were identified as riskier. To favor visualization and integration with commonly used GIS platform, results were exported in KML (Google Earth) and SHP (ArcGIS) formats.

The authors believe this approach can be implemented as a map-production workflow where routine monitoring of satellite data is pushed within a GIS-integrated analysis pipeline to be analyzed by a set of plugins designed to monitor/detect potentially hazardous features, and the results exported as Google Earth (KML) files or ArcGIS layers to provide immediate visualization and delivery. Other geospatial data layers, such as geology, karst, soils maps, or fault zones, can be delivered on the same platforms, thus offering greater efficiency and geospatial data integration in planning, inspection, and incident response.

Disclaimer: The views, opinions, findings and conclusions reflected in this paper are the responsibility of the authors only and do not represent the official policy or position of the US DOT/RITA, or any State or other entity.