Critical Infrastructure Resilience

With an emphasis on outputs-oriented research, education and workforce development, and continuous engagement with communities and practitioners, our research in this area has focused on modeling the vulnerability of interdependent critical infrastructure systems to natural hazards, improving equitable resilience of infrastructures and communities, understanding the impact of prescribed burns on wildfire reduction and stakeholder values, and designing optimal resource allocations integrating human behaviors for rapid recovery and reconstruction of communities in the aftermath of a disaster.

This work has been funded by various programs of the National Science Foundation, such as the Humans, Disasters and the Built Environment (HDBE) and the Strengthening American Infrastructure (SAI), the Natural Hazards Center, UB SUNY internal SEED grants, UB Center for Geohazards Study, etc.

Wildfires represent a significant threat to critical infrastructure systems. These challenges are compounded by the observation that rural and disadvantaged communities are often the most susceptible to wildfire disasters. Although significant progress has been made in predicting wildfire propagation, less is known about the complex interactions between wildfires, socially vulnerable populations, and emergency management practices. This project focuses on strengthening the emergency management of critical infrastructure systems, with particular attention to the disproportionate societal impacts of wildfires. The goal is to develop a new assessment framework and an integrative decision model that enhances the human-centered governance of such systems. Specifically, this project integrates social scientific theories with mathematical models to yield novel insights into the design and improvement of emergency management of critical infrastructure systems to develop a human-centered, equity-focused, risk-informed decision-making framework to address these challenges. The research develops equity-aware and interpretable models and computational algorithms for vulnerability assessment and efficient post-wildfire recovery strategies of interdependent critical infrastructure systems under deep uncertainties.

Wildland fires, exacerbated by climate change, pose a significant threat to communities, resulting in deaths, injuries, and substantial economic costs. Prescribed burning, increasingly used to mitigate these fires, faces challenges such as funding shortages and conflicting stakeholder objectives. This project aims to quantify the effectiveness of prescribed burning in reducing wildland fires, using new datasets and models to enhance burn planning and reduce losses in high-risk areas. It will involve collecting and analyzing comprehensive data on fire occurrences, prescribed burn usage, stakeholder objectives, and environmental factors to develop analytical models that assess the impact of prescribed burns on fire reduction and stakeholder values.

Hurricane Fiona severely impacted Southwestern Puerto Rico, provoking landslides, unleashing flooding across the Island, and obliterating the power grid. Such impacts are unusual for a Category 1 hurricane, raising concerns about the disaster-preparedness of the US Territory that faces a multilayered socioeconomic crisis rooted in long-standing policy, migration, and poor budget practices. While it is well-known that disasters and subsequent recovery efforts exacerbate the socioeconomic disparities among marginalized groups, there is a lack of understanding of how these groups are being impacted by the cascading critical infrastructure (CI) failures and delays in the restoration process. This research studies the impacts of cascading failures of power grid and transportation systems on socially vulnerable communities. Through fieldwork data, it understands the dynamic relationships of CI restoration interdependencies with the community recovery process. The research team conducted stakeholder interviews, community focus group interviews, and surveys, and used a mixed-methods approach to conduct the analysis. In addition, the team analyzed data extracted newspaper articles using Artificial Intelligence approaches, processed and cleaned publicly-available data from GIS, Census Bureau, etc.

Unequal access to essential services (e.g., healthcare, food) is a pressing issue impacting communities worldwide. Particularly, socially vulnerable populations often experience more significant disparities in service access, leading to the emergence of healthcare deserts and food deserts. Previous studies mainly use static measures (e.g., proximity and availability) to understand people’s potential access to services, overlooking the dynamic aspect of human mobility, which is crucial to accurately reflect the actual access of people. To address this research gap, this project focuses on characterizing, predicting, and improving people’s access to essential services, leveraging large-scale mobile phone data. Various modeling approaches this research uses include bipartite network development and analysis, advanced multivariate time series analysis, optimization, and risk-informed resource allocation decision-making.

Recent natural disasters provide evidence of increasing recurrence and disruptive forces of extreme events and reveal the extent of vulnerability of our energy systems in the face of such events. Besides climate change and extreme weather events, population growth, accelerated urbanism, aging infrastructure, chronic underinvestment in infrastructure maintenance/rehabilitation, and governance issues are key stressors that threaten the proper functionality of the energy sector in the U.S. Faculty research focuses on addressing the various challenges in attaining climate-resilient, sustainable energy infrastructure. This research focuses on modeling the impacts of climate change, such as increased temperature and heatwaves on the nation’s energy system, quantifying the effects of natural hazards, such as hurricanes and wildfires, on the electric grid, and improving the equitable resilience of the nation’s grid under the threats of a compound disaster such as the co-occurring wildfire and heatwaves, to inform resilience building decisions equitably. Various social-science-informed mathematical models using data-driven techniques such as advanced machine learning, artificial Intelligence, spatiotemporal analysis, and optimization are leveraged to address the pressing challenges in this research domain.

Healthcare access is critical to every society and necessary to maintain the well-being of its inhabitants. Yet, the healthcare system might not be completely functional after a disaster due to several factors. Health care facilities might not be open due to structural damage of their facilities, communication and electrical disruptions, and lack of medical supplies. Access may also be affected by the roads, highways, and bridges that impede the transportation of individuals to the health facilities. Yet, the transportation system is often overlooked by the public health perspective. This research investigates the healthcare and transportation barriers present in socially vulnerable communities after a disaster in Puerto Rico. Data from the Centers for Disease Control and Prevention (CDC)’s Social Vulnerability Index at the county-level is analyzed to measure social vulnerability in Puerto Rico. Other key variables include transportation network data, i.e., distances from the municipalities’ center to the nearest key facilities, which are used to measure the transportation barriers. The analysis also uses health variables like number of hospitals, percent of population with Medicare/Medicaid, State, and Children’s Insurance. This study employs principal component analysis and ordinary least squares regression to investigate the correlation between transportation networks, healthcare variables, and social vulnerability. These findings, coupled with interview data, help explain how transportation access, including network vulnerabilities, poses a challenge to the most socially vulnerable communities in post-disaster environments.