Topological network and GIS approach to modeling earthquake risk of infrastructure systems: A case study in Japan

Abstract

In this study, we propose a network-geographic information system (GIS) modeling framework to construct a risk map for revealing infrastructure vulnerability to seismic hazards. The framework considers the cascading network of interdependencies among the effects and intermediate effects of earthquakes. Network modeling is applied to analyze infrastructure vulnerabilities and identify the infrastructure types that are vulnerable to earthquakes, while geographical modeling is applied to identify seismic hazards in terms of ground motion intensity and earthquake source faults. Reported earthquake cases in Japan are selected for analysis, and an illustrative example of the modeling of a risk map is conducted on a section of a conventional train line in Japan. In the network modeling, statistics and network diagrams are developed to analyze the relational earthquake incident data. The results demonstrate that earthquakes can cause direct and indirect effects on infrastructure, and interdependencies can be found among the effects and intermediate effects. In the geographical modeling, probabilistic distribution is adopted to predict the seismic characteristics; the results reveal the intensity occurrence probability and cascading consequences of different earthquake intensities, and predict specific degrees of destruction. Infrastructure vulnerability and seismic hazard maps are developed according to the analyzed results. A risk map is then constructed by integrating the infrastructure vulnerability map with the seismic hazard map. This study characterizes the risk of earthquakes to infrastructures, capturing the complex interdependencies to reveal the cascading effects from earthquakes, and quantifying and demonstrating infrastructure areas vulnerable to seismic hazards.