ANOMALOUS DISPLACEMENT DETECTION OF BRIDGES USING SATELLITE SAR IMAGES: A CASE STUDY ON MUSOTA WATER PIPE BRIDGE IN WAKAYAMA

Abstract

This paper analyzes a displacement of Musota water pipe bridge in Wakayama, where a collapse accident occurred, using satellite SAR images, and discusses the possibility of anomaly detection. First, the results of the displacement analysis show that six of seven spans which did not collapse have similar displacement trends, while the collapsed span has anomalous displacement that seemed to be a sign of collapse about one year before the accident. Second, the results of the correlation analysis between displacement and environmental factors (atmospheric temperature and wind speed data) show a weak correlation with atmospheric temperature, while a negative correlation with wind speed. Finally, this paper proposes a clustering method for detecting "a span with anomalous displacement" using the displacement data, and show that it is possible to detect anomalies on the collapsed span based on objective indicators.