PSI - Issue 64

H.F. Stewart et al. / Procedia Structural Integrity 64 (2024) 573–579 Stewart, Cusson, Greene Gondi & Oliver / Structural Integrity Procedia 00 (2024) 000–000

574

2

1. Introduction 1.1. Background

Canada is a geographically large country with an aging public infrastructure. Several bridges are considered deficient in structural capacity or functionality, or are approaching the end of their design service life (FCM, 2019). These bridges may be separated by large distances or located in remote areas, creating logistical challenges for timely in-situ inspections. Satellite remote sensing (SRS) offers bridge operators a means to monitor the health of multiple structures over large geographic areas without in-situ instrumentation. To this end, the National Research Council of Canada (NRC) is overseeing the development of a Remote Sensing Structural Health Monitoring (RS-SHM) method that allows transportation authorities to monitor the behaviour of bridges and other civil structures over time using satellite-based radar, multispectral, and thermal imagery. This RS-SHM method supplements and complements in-situ visual inspections and provides bridge operators with a means of assessing structural performance between periodic inspections. It can also act as an early detection and warning system for emerging structural health problems or environmental changes that may pose an elevated level of risk to a given structure or group of structures. Persistent Scatterer Interferometry (PSI) is the backbone of the RS-SHM method. This technique uses repeated acquisition of Synthetic Aperture Radar (SAR) images from the same satellite sensor and viewing geometry over time to determine the displacement of repeatedly observable coherent target points (persistent scatterers, or PS) on the structure. With sufficiently many PS targets, it is possible to produce a dense geodetic network over a structure and observe spatial and temporal deformations at small scales, which can be compared to theoretical estimates and field measurements of structural behaviour. Concurrently acquired high-resolution multispectral and thermal infrared satellite imagery and in-situ time-series environmental data can provide context for interpreting PS-InSAR measurements (Cusson and Stewart, 2024). Accurate geolocation of PS targets’ horizontal positions and elevations is necessary for reliable PSI deformation results. During a validation trial of the RS-SHM method on the Samuel de Champlain Bridge in Montreal (QC), Canada, we observed that a lack of adequate knowledge of the height of elevated bridge components contributed to geolocation errors of PS-InSAR targets, since the bridge construction drawings were not available to the research team when the InSAR analysis was conducted. Such geolocation errors may lead to erroneous interpretation of the data, which in turn could compromise the effectiveness of the RS-SHM method. While bridge drawings, Lidar surveys, or dimensional control surveys are preferred for determining elevations above a project height datum, these drawings or surveys may be unavailable, unreliable, or obsolete (such as original as-built drawing elevations after decades of known ground subsidence). A Digital Surface Model (DSM) from satellite stereo imagery is a remote sensing technology that can provide height information side-by-side with contextual information from the source images. It is our goal in this paper to evaluate whether a DSM derived from an Airbus Pleiades tri-stereo image triplet is a suitable source of height corrections for adequately geolocating PS-InSAR measurements on bridges if no other higher-order source of height data was available. 1.2. Study area The Samuel de Champlain Bridge (SdC) is a multi-transportation-mode cable-stayed bridge crossing the Saint Lawrence River in Montreal (QC), Canada, that opened to vehicle traffic in July 2019. This new bridge replaced the old Champlain Bridge after 57 years of service. Fig. 1 shows the Saint Lawrence River in metropolitan Montreal. Bridges and structures discussed in this study are labelled, and an inset shows a digital rendering of the Samuel de Champlain Bridge. The Saint Lawrence River in Montreal is separated by the Saint Lawrence Seaway Levee into the fast-flowing, variable water level Greater La Prairie Basin to the west and the lock-and-levee controlled Lesser La Prairie Basin and Seaway navigation channel to the east. Oceangoing vessels with air drafts of up to 35.5 meters transit through the Seaway, and statutory requirements on safe vessel navigation specify that the minimum vertical clearance of the Main Span’s deck bottom must be at least 36.6 meters (120 feet) above the Seaway high water level (SLSMC, 2019). The distance from the East Abutment (EA) of the SdC Bridge to the main span tower (1086 m) is less than half the distance from the West Abutment (WA) to the tower (2249 m), creating an asymmetrical elevation profile of the bridge.