Critical infrastructure networks in flood risk management : integration of flood risk analysis methods for networks, consideration of network-specific measures and challenges for data availability

Critical infrastructure (CI) networks are vital for the survival and functionality of society and economy. In an environment with a changing climate and more extreme weather events, it is pivotal to prepare infrastructure for these changes. Flooding, one of the major natural hazards that societies are exposed to, is addressed by constantly evolving flood risk management (FRM). However, disruptions to CI services and the cascading effects of these disruptions are not currently included in FRM. In flood risk analysis, it is state-of-the-art to determine the direct consequences of flooding on assets and people. Disrupted CI services and their networks, as well as the cascading effects propagating flood disruptions, are not considered explicitly, and methods to derive these consequences are scarce. To address these challenges, a cumulative dissertation consisting of four manuscripts was prepared. The research presented in the first manuscript of this dissertation incorporated CI into every facet of FRM, encompassing flood risk analysis, risk reduction as well as risk communication and participation. Utilising a CI network modelling technique, this approach facilitates the quantification of flood consequences for the CI within flood risk analysis. Consequently, the total flood risk, as a combination of hydrological probabilities and associated consequences, is complemented by CI consequences. The integration of CI and its network characteristics allows for enhanced decision-making in risk management. Moreover, the ongoing involvement of CI operators has proven to be advantageous throughout all FRM stages. A case study conducted in Accra, Ghana, including the participatory engagement of CI stakeholders, underlines the previously mentioned findings and the continuous integration of CI into FRM. The flood risk analysis, as a component of FRM, relies on the consideration of consequences specific to CI networks; therefore, a modelling approach of CI networks is designed. The modelling approach was conceptualised and implemented to balance simplicity and accuracy to represent the complexity of CI. In the second manuscript, catchment-wide flood risk analyses utilising a topology-based modelling approach for CI networks are proposed. The fundamental elements of the model encompass points, connectors, and polygons, which are employed to depict the layers and multisectorality of the CI network. This newly formulated approach was technically integrated as a CI network module into the PROMAIDES (Protection Measures against Inundation Decision support) framework for more integrated flood risk management. This method quantifies the consequences of supplying the number of affected CI users in combination with the duration of disruption. A proof-of-concept of this modelling approach was conducted in the case study area of Accra, Ghana. FRM requires the consideration of risk management measures by definition and, in the context of this work, must account for measures within the CI domain. Many CI operators have valuable experience in managing flood risk within their sectors, offering insights into potentially effective flood risk reduction measures. The third manuscript presented in this dissertation conducts structured interviews with CI operators in Central Europe supplemented by a preparatory and follow-up literature review. The findings from the combined literature review and expert interviews were compiled into a comprehensive catalogue detailing flood risk reduction measures tailored to five CI sectors as well as a generalised measure category. A case study in Western Germany is introduced, illustrating the integration of five CI-specific flood measures from the catalogue into the flood risk analysis and management framework. A persistent challenge across CI network modelling approaches is the lack of high-quality input and validation data. Consequently, modellers often resort to assumptions or extrapolations that introduce uncertainties that undermine model reliability. To address this challenge, a generic modelling workflow is proposed in the fourth manuscript, and the data requirements for each stage are systematically defined, with a review of potential sources to enhance data collection and awareness of pitfalls. Case studies illustrate the application of this workflow across different modelling purposes, emphasising the need to explore the implications of data scarcity and improve the reliability of the CI network model. This dissertation advances the enhancement of critical infrastructure resilience during flood events. For the first time, it demonstrates that integrating critical infrastructure into flood risk management is not only feasible but has been successfully achieved. Special attention is given to the analysis of flood risk for CI networks because of the individual composition of these networks, as well as to the cascading effects that propagate failure due to flooding through the CI networks to other elements and sectors. Next, it is concluded that CI-specific measures are necessary, as collected in a CI-specific flood risk reduction measure catalogue. Estimating CI consequences for decision-making hinges on solid data inputs and can suffer when these inputs are scarce. However, it has been shown that raising awareness of the necessary requirements, engaging CI stakeholders, and ensuring proper distribution can overcome this shortfall.