UC Berkeley

Critical infrastructure networks provide socio-economic services across local and regional scales. However, population and economic growth, urbanization, increasing infrastructure interdependency, and natural hazards – not least due to the worsening effects of climate change – are putting these infrastructure systems under increasing pressure. A comprehensive understanding of the potential impact of climate-change-induced natural hazards on the intra- and interconnected critical infrastructure networks and network resilience options given future climatic variations is still absent. Through three types of infrastructure networks as case studies, this dissertation develops a holistic and proactive approach to gain a better understanding of climate change impacts on critical infrastructure networks. The first case study focuses on the multimodal fuel transportation network in the San Francisco Bay Area and examines the local and cascading impact of coastal flooding under 120 climate change scenarios (including four global climate models, two representative concentration pathways, three sea level rise percentile estimates, and five planning horizons). These scenarios cover a wide range of possible climate outcomes from the most conservative to the most radical. The known extremes in current climate change predictions are taken into consideration. A multimodal network model is built to capture the connectivity and flow between different infrastructure assets. We leveraged outputs from state-of-the-art hydrodynamic flood models and evaluated the direct exposure of physical infrastructure exposure and indirect cascading network impacts. Our results reveal increasing fragmentation, a decrease in network density, connectivity, and efficiency from the year 2000 to the end of the century. When taking cascading effects into consideration, the actual impact of flood hazard is greater than that estimated from exposure analysis. Building on the first, the second case study explores further into hazard impact patterns using innovative pattern recognition methods. Using the National Airspace System (NAS) infrastructure as an example, we develop a new approach to understanding the spatiotemporal impact patterns of lightning strike hazards, which is crucial for guiding the design of a resilient NAS. We discover a total of five unique hazard impact patterns leveraging 11 years of lightning occurrence data and outage data from the Federal Aviation Administration. Among them, four impact patterns reveal new spatiotemporal characteristics of the lightning hazard that are absent in the current aviation knowledge base and not yet incorporated in existing lightning protection policies and procedures. Finally, we explore beyond local, regional, and national cases and aim to find universal network resilience patterns in face of climate change impacts. In the last study, we provide the first global evaluation of urban road networks in terms of both direct exposure to flood hazards and indirect impacts due to city-wide travel disruptions and cascading failures. We create a dataset of topological road networks for 2,564 cities in 177 countries, covering over 14 million kilometers of roads, and considered ten probabilistic flood scenarios (1:5 year to 1:1000-year return periods). Our results show heterogeneous flood impacts on road infrastructure and mobility across different countries and regions. It enables comparisons of exposure and vulnerability of road networks to flood hazards across countries, allowing the identification and prioritization of urban transport resilience measures. Furthermore, it reveals patterns in network topology and flood exposure that influence the overall resilience of networks, providing a foundation for causal factor identification. The three studies create a cohesive narrative about critical infrastructure resilience in face of natural hazards under a changing climate and offer insights into a new paradigm for environmental planning. Collectively they develop new strategies for detecting exposure, vulnerability, and resilience in intra and interconnected infrastructures. This new paradigm encompasses the complexity of the interrelated human and natural environment and is necessary if we are to predict and deal with our future habitats under climate change.