Stress-testing global infrastructure networks to synchronous flood events

Civil infrastructures, such as buildings, roads and electricity networks, form the backbone of society and support human wellbeing (Thacker, et al., 2019). Given society's strong dependence on infrastructure systems, disruptions impose a heavy economic burden globally, with floods accounting for $1,092 billion in losses since 1980 (Munich Re, 2019). Ensuring infrastructure systems are resilient to shocks is becoming an increasingly urgent task in the context of growing populations and heightened likelihoods of extreme weather due to climate change (Winsemius, et al., 2016; Forzieri, et al., 2018; Hall, et al., 2019). However, existing large-scale studies that focus on infrastructure damages caused by flooding (Hirabayashi, 2013; Winsemius, et al., 2013; Koks, et al., 2019b) are severely limited by two implicit assumptions: 1) that flood events occur independently from each other across space; and 2) that consequences of one asset failing is isolated to that single asset, without accounting for network effects1. This has major implications for our understanding of what the worst-case hazard scenarios are for infrastructure systems; for example, simultaneous floods may induce widespread losses that overwhelm insurance claims (Jongman, et al., 2014). In order to systematically identify vulnerable assets and target adaptation funds, infrastructure systems
should be stress-tested2 against a large ensemble of realistic synchronous flood events.