Timescale: Oct. 2021 – Sept. 2024
PD Dr. Norbert Marwan, PIK Potsdam & University of Potsdam
Prof. Bruno Merz, GFZ Potsdam & University of Potsdam
Extreme weather events can affect significantly infrastructure networks, e.g., flooding can disrupt transportation networks, storm gusts can distort communication networks and cause power blackouts. The increasing interdependencies between the different infrastructure sectors can result in cascading failures. Changing climate and changing infrastructure topology as well as growing interdependencies between the sectors rise concerns about multiplying risks and require effective adaptation and mitigation strategies. However, traditional impact and risk analyses are limited when considering the complex and coupled nature of the infrastructure sectors.
Complex networks have been proposed and applied to investigate impacts and mechanisms, and to suggest countermeasures for mitigating the impact of disruptive events on infrastructure networks (e.g., Pregnolato et al. 2017; Fu et al. 2017). The interdependency network model covers three types of interdependencies: extent, directionality, and redundancy (Fu et al. 2014). These three factors play different roles in cascading propagation of failures in coupled networks and allow different strategies of mitigation. Moreover, it has been demonstrated that traditional protection strategies, such as securing network hubs or isolating affected network parts, are less effective for interdependency networks (Gao et al. 2011; Khoury et al. 2015). Thus, novel approaches for estimating the vulnerability of networks and coupled networks have been suggested in the last years, such as the ability of rewiring in engineered interdependent networks (Khoury et al. 2015).
This project will further develop and apply the approach of interdependency networks on infrastructure data in order to evaluate the impact of extreme rainfall and flood events. The focus will be on the evolution of the topology of the interdependency networks (using past data and projections) of transportation networks and energy networks (gas, power grids). Changes and transitions in the extent, directionality, and redundancy in the networks will be modelled using conceptual network models and estimated from real infrastructure networks. The impact of local and large-scale disruptive events will be studied for different strongly connected networks. Scenarios for disruptive events will be derived from past observations and from a flood modelling chain. The project will further study the potential changes in the severity of climate-related extreme rainfall and flood events due to potential changes in the occurrence and synchronization of extreme events (derived from climate projections and the PhD projects Q2 and Q6).
The investigation of robustness and vulnerability of interdependency networks during disruptive (extreme weather) events will provide new insights to the changing risk of infrastructure failures and provide measures for improving the resilience of such infrastructure networks. The work will result in new methodological developments in the field of network of networks and vulnerability analysis using network topology statistics.
Dedicated Regional Cluster: Central Europe / Germany
Responsibilities: The PhD-project “Impact of extreme events on topological robustness of interdependent infrastructure networks” is based at the research department “Complexity Science” at PIK Potsdam. The aim of the project is to investigate several aspects of robustness and vulnerability of interdependent infrastructure networks during local and large-scale disruptive (extreme weather) events and due to potential changes in the occurrence and synchronization of such extreme events. Changes and transitions in the extent, directionality, and redundancy in the networks will be modelled using conceptual network models and estimated from real infrastructure networks, whereas disruptive events will be derived from past observations and from a flood modelling chain. The study will further attempt to develop measures for improving the resilience of the infrastructure networks. Methods for this project will involve networks of networks, recurrence analysis, and Bayesian statistics.
Requirements: We are seeking applications from highly motivated individuals with a strong background in physics or mathematics, statistics, and data analysis as well as very good knowledge and experience in programming (preferably Python or Julia). Knowledge on hydro-meteorological processes is desirable. Fluency in the English language (speaking and writing) as well as the willingness to work in an interdisciplinary team are essential.