Physical infrastructure interdependency and regional resilience index after the 2011 Tohoku Earthquake in Japan

A resilience index is used to quantify preventive measures, emergency measures, and restoration measures of complex systems, such as physical infrastructures, when they are subjected to natural disasters like earthquakes, hurricanes, floods, etc. Interdependencies among these systems can generate cascading failures or amplification effects, which can also affect the restoration measures right after an extreme event and generate a reduction of the resilience index. In this article, a method is proposed to evaluate the physical infrastructure resilience of a region affected by a disaster considering infrastructure interdependency. It is illustrated using available restoration curves from the March 11 2011 Tohoku Earthquake in Japan. The weights assigned to each infrastructure, which are used to determine resilience, are evaluated using the degree of interdependency indices which are obtain by time series analysis. Results show that the weight coefficients thus obtained do not influence the resilience index significantly; however, the methodology proposed is unbiased from subjective judgment and is able to identify the critical lifelines. Furthermore, the results of the case study presented here suggest that to obtain meaningful estimation of the weight coefficients, it is necessary to consider the period range between two perturbations (e.g., main shock and aftershock). Future infrastructure disruption data (from this and other earthquakes) would be needed to generalize this finding that will allow also to quantify the changes in the restoration curves caused by the magnitude and distance of the shocks from the epicenter, as well as the intrinsic properties of the physical infrastructures. Copyright © 2014 John Wiley & Sons, Ltd.