Abstract
This research aims to investigate a method for estimating the production capacity loss rate (PCLR) of industrial sectors damaged by a disaster, such as an earthquake, tsunami, or nuclear radiation, particularly the 2011 Great East Japan Earthquake. PCLR is fundamental information required to gain an understanding of economic losses caused by a disaster. In particular, this paper proposes a method of PCLR estimation that considers the two main causes of capacity losses as observed from past earthquake disasters, namely damage to production facilities and disruption of lifeline systems. To achieve the quantitative estimation of PCLR, functional fragility curves considering the relationship between production capacity and earthquake ground motion and lifeline resilience factors for capturing the impact of lifeline disruptions have been adopted, while actual recovery curves are considered mainly for damaged facilities. Through the application of this method to the case study of the 2011 Great East Japan Earthquake, the PCLR in various industrial sectors is estimated; the estimated PCLR in the manufacturing sectors are then compared to the corresponding index of industrial production. The results demonstrate that the estimated values are close to the actual production indices in the overall manufacturing sector and many of the individual sectors.
Acknowledgments
We are grateful to editors and anonymous reviewers who made valuable comments, significantly improving the quality of the paper. We also acknowledge that the research was carried out with the partial supports of JSPS KAKENHI Grant Number 24710185 and the Program for Risk Information on Climate Change funded by the MEXT-Japan.
Notes
1 In Mackenzie et al. (Citation2012), direct impacts are determined as production decrease due to the destroyed or partly damaged facilities and indirect impacts are production losses due to the intermediate and final demand changes.
2 In fact, more prefectures suffered direct damages, including Chiba, Aomori and Nagano prefectures. These prefectures should also be analyzed in a similar manner.
3 In other words, uniform distribution is applied to the range in each category. For example, because functional damage levels 1/4 and 1/5 belong to the same category I, the occurrence probabilities of these are same. The number of categories becomes larger, and more detailed results can be obtained. Due to the limitation of sample number for the functional fragility curve, the three categories are used for now.