Application of High-Resolution Optical Satellite Imagery for Post-Earthquake Damage Assessment: The 2003 Boumerdes (Algeria) and Bam (Iran) Earthquakes

Remote sensing technology is increasingly recognized as a valuablepost-earthquake damage assessment tool. Recent studies performed by research teams in the U.S., Japan and Europe have demonstrated that building damage sustained in urban environments can be readily identified through the analysis of optical (Matsuoka and Yamazaki, 1998; Chiroiu et al., 2002; Huyck et al., 2002; Mitomi et al., 2002; Yusuf et al., 2002; Shinozukaet al., 2000; Saito and Spence, 2004) and synthetic aperature radar (SAR) (Aoki et al., 1998, Huyck et al., 2002; Yusuf et al., 2002) imagery. Under the broad aim of identifying ways in which post-earthquake response and recovery activities can be improved through the integration of remote sensing technologies, a Multidisciplinary Center for Earthquake Engineering Research (MCEER) team has been investigating their use for urban damage detection and emergency response (Eguchi et al., 1999; Eguchi et al., 2000a, 2000b; Eguchi et al., 2003; Huyck and Adams, 2002;Huyck et al., 2002). Research to date has focused on various aspects of damage detection, including the development of post-earthquake damage detection algorithms that use optical and SAR data to locate building collapse, and a mapping system to display and disseminate earthquake-related multimodal geospatial data. This paper describes an extension of the methodology previously developed for the 1999 Marmara (Turkey) earthquake (see Eguchi et al., 2003), to detect building damage caused by the 2003 Boumerdes (Algeria) and Bam (Iran) earthquakes (see also Adams et al., 2003a, 2003b, 2004). For the Marmara event, change detection algorithms were based on moderate resolution SPOT4 optical and ERS SAR coverage. Following the recent launch of Quickbird and IKONOS commercial satellite systems, a new generation of very high-resolution imagery has become available. The Boumerdes 174 Emergency Response and Recovery and Bam earthquakes mark the first two occasions for which this imagery was collected before and soon after the event. However, the increase in spatial resolution from10 meter to sub-meter accuracy calls for a number of methodological refinements. Initially, a ‘Tiered Reconnaissance System’ (TRS) is conceptualized, which extends the scope of information collected to include damage visualization for individual structures. Following details of a methodological refinement at the pre-processing stage, implementation of the revised methodology is described for the Boumerdes and Bam events. The paper goes on to provide detailsof its subsequent deployment asa post-earthquake reconnaissance tool within the VIEWS (Visualizing Impacts of Earthquake With Satellites) system, before concluding with a brief discussion of future research directions.