InSAR.no: A National InSAR Deformation Mapping Service
John Dehls is a research scientist at the Geological Survey of Norway (NGU). He has an MSc from Lakehead University (1992) and a PhD from University of Toronto (1997). His last fifteen years have been focused on radar remote sensing; specifically the use of interferometric synthetic aperture radar (InSAR) techniques for deformation monitoring. He has led the development of an operative landslide mapping and monitoring system using InSAR within NGU based upon Radarsat-2 data, and is currently leading the establishment of a national InSAR service based upon data from the Copernicus Programme's Sentinel-1 satellites.
Systematic mapping of unstable rock slopes has been carried out in Norway since 2005. More than 300 unstable or potential unstable rock slopes have been detected and characterized so far. This requires a standardized hazard and risk classification system, which we established in 2012. The determination of the hazard and risk level follows a new standard approach for the systematic mapping of potentially unstable sites. The mapping approach has been designed to focus effort on sites with higher levels of risk. Rock slope failures that would not result in any loss of life, as there is no life line or building in the run-out area, are mapped without a probability assessment and thus not assigned a hazard class. Rock slope failures that can result in loss of life are analysed using qualitative hazard analyses, thus mapping products are hazard maps with qualitative probability classes.
Both the systematic mapping and the subsequent hazard and risk classification are dependent upon knowledge of deformation rates and possible acceleration. Given the large size of Norway and the number of sites that are being investigated, we are reliant on remote sensing techniques to provide most of this information. Interferometric synthetic aperture radar (InSAR) is a technique that is able to measure displacements over large areas with an accuracy of better than one millimetre per year. We have been systematically collecting Radarsat-2 data over much of Norway since 2009 for identifying and periodically monitoring unstable rock slopes.
In May 2016, we launched a project to develop a national InSAR-based deformation mapping service, based upon the Sentinel1 satellites. Its mandate is to provide the public in Norway with nationwide deformation products. The service will provide periodically updated deformation data, with varying resolution for urban and non-urban areas. The products will be made available to various local, regional and national authorities via appropriate web GIS protocols. The data will also be made available to the public via a web map interface with simple tools to query and visualize the information.
Scaling up from regional operations, based upon data every 24 days, to a national operation, with data every 6 days, is challenging. In order to process the huge quantities of images being received, we are moving from manual, interactive processing of individual data stacks, to automated ingestion of data with dynamic updates of deformation estimates. In this talk I will present examples of results from both Radarsat-2 and Sentinel-1, with a focus on application to landslide hazard mapping, but will also touch upon urban applications.