Celebrating successful earthquake risk reduction through counterfactual probabilistic analysis

• Main Authors: Rabonza, Maricar, Lin, Yolanda, Lallemant, David
• Other Authors: Asian School of the Environment
• Format: Conference or Workshop Item
• Language: English
• Published: 2021
• Subjects: Engineering::Computer science and engineering::Mathematics of computing::Numerical analysis; Engineering::Computer science and engineering::Mathematics of computing::Probability and statistics; Social sciences::Geography::Natural disasters; Earthquake Risk Assessment; Probabilistic Analysis; Counterfactuals
• Online Access: https://hdl.handle.net/10356/152837
• Institution: Nanyang Technological University

While thousands of lives are saved through effective earthquake risk reduction programs, very few opportunities exist to celebrate such successes. First, if the tangible benefits of such interventions are only felt in the advent of a large earthquake, there is often a large time-delay between the intervention and its benefits. Second, large earthquakes are rarely an appropriate time for celebration since even small losses cause suffering, and the argument that losses could have been worse had it not been for a successful risk reduction program is not particularly heartening for those who have nevertheless suffered. Therefore, we are calling for a more systematic assessment of probabilistic lives saved (or avoided losses) from earthquake risk reduction actions, and celebration of these successes before risk reduction programs demonstrate their benefits. To address this challenge, this work presents a stochastic framework to estimate the mitigation effect of large-scale earthquake risk reduction measures in terms of probabilistic lives saved. The framework implements a counterfactual approach by analysing the probabilistic consequences of an earthquake had a certain risk reduction program not been implemented beforehand. Two main applications are presented: (1) Calculating the benefits of an earthquake risk mitigation action in an actual past earthquake, and (2) calculating probabilistic benefits over the lifetime of an intervention. For the former, probabilistic realizations of earthquake casualties without the risk intervention are modelled and compared with actual losses - a powerful representation of the success of a risk intervention. For the latter, probabilistic lives saved are calculated based on the hazard model rather than an actual past event. Since future risk is dynamic, we further make use of recently developed time-dependent exposure and vulnerability models to study longterm risk. Ultimately, this study demonstrates the use of counterfactual probabilistic risk analysis as a method for assessing probabilistic benefits of disaster risk reduction actions, so that they may be celebrated. We hope that such methods could someday lead to a "Probabilistic Lives Saved Award," newspaper headlines highlighting disaster risk reduction successes, and ultimately reinforce more investments in such earthquake risk reduction policies and actions.