The resilience of a community depends on the performance of the built environment (e.g., building portfolios and infrastructure systems) as well as the functionality of its constituting social and economic institutions. Communities are expected to be resilient to disruptive events. That is, they should possess the ability of limited disruption, immediate response, and quick recovery [1,2]. Remarkably, the 2019 coronavirus disease (COVID-19) epidemic [3] has resulted in a significant public health emergency around the world, raising unprecedented challenges to the modern communities. Furthermore, in recent years, other types of disruptive events, such as earthquakes, tropical cyclones, floods, and wildfires, have threatened communities globally with dramatic consequences. With respect to this, the society and the public are asking justified questions such as the following: How resilient is our community to disruptive events? How can we use resilience approaches to counteract disruptive events? What lessons can we learn from real-world practices to enhance the resilience of our community?
The focus of this Special Section is on the methods, lessons learned, and developed practices that are helpful to achieve community resilience in the face of disruptive events, including earthquakes, tropical cyclones (hurricanes), and sanitary crises such as the COVID-19 pandemic. This Special Section contains a collection of six research papers, as highlighted herein.
Capshaw and Padgett developed a predictive model for the likelihood and expected duration of refinery shutdowns under hurricane hazards. This model was applied to predict the U.S. Gulf Coast refinery downtime using approximately 120 historical observations from storms impacting U.S. Gulf Coast refining centers over the past 20 years. The authors demonstrated the multiscale and cascading consequences that storm-induced facility impacts can have on local to global communities and their markets.
Trump, Jin, Galaitsi, Cummings, Jarman, Greer, Sharma, and Linkov developed a three-part data visualization methodology to assess COVID-19 vaccination equity in the United States using state health department, U.S. Census, and Centers for Disease Control and Prevention data. The three parts include: (1) the equitable pathway deviation index, (2) assessment of perceived access and public intentions to vaccinate, and (3) equity derivative in social vulnerability analysis. The authors also emphasized the importance of long-term investment in developing the necessary expertise and mission to implement equity mandates.
To overcome the challenge of obtaining sufficient operation and failure data to train machine learning (ML) models in pipeline failure risk prediction, Mazumder, Modanwal, and Li employed a generative adversarial network-based framework to generate synthetic pipeline data using a subset of experimental burst test results data compiled from the literature. They also trained the random forest models to investigate the efficiency of ML on real and synthetic data for predicting the failure of oil and gas pipelines. The authors concluded that with the incorporation of synthetic data, the random forest model can predict burst failure risk classifications with improved accuracy than using real data only.
Kwon and Song applied β−π analysis (β = reliability index and π = redundancy index) to the disaster resilience evaluation of infrastructure networks from a view of system reliability. They also proposed a normalized causality-based importance measure to feature a component's relative importance through network topology and reliability. The authors utilized a numerical example of a bridge network to demonstrate the applicability of the proposed system-reliability-based disaster resilience analysis framework, finding that the normalized causality-based importance measure is able to achieve a balanced consideration of a component's topological location in the network and the component reliability.
Wang, Ayyub, and Beer explored the relationship between time-invariant reliability-, time-invariant resilience-, time-dependent reliability-, and time-dependent resilience-based design methods. They also developed a load and resistance factor design-like design criterion for structural resilience-based design, namely, load and resilience capacity factor design. This is based on a new concept of structural resilience capacity, which is a generalization of structural load bearing capacity (resistance). The two criteria—load and resistance factor design and load and resilience capacity factor design—can be used together simultaneously to meet reliability and resilience goals of the designed structure.
Montoya-Rincon, Gonzalez-Cruz, and Jensen employed a machine learning approach to evaluate the effects of strengthening scenarios of power transmission lines. They analyzed three direct hardening scenarios and determined the effectiveness of the changes in the context of Hurricane Maria (2017). It is found that steel self-support poles are an ideal replacement option for the wooden poles, reducing the damaged structures in the section of the line by 40%. The authors also concluded that all the three hardening scenarios are a viable option to increase the resiliency of the transmission lines.
Finally, the guest editors would like to express their sincere gratitude to the authors and reviewers who have contributed significantly to the preparation of this Special Section. It is hoped that these papers will be beneficial for readers to deepen their understanding of community resilience.
