A Network Analysis Approach to Infrastructure Resiliency to Compounding and Nonstationary Threats

Urban infrastructure systems are vulnerable to both anthropogenic and natural threats that have inherent uncertainty in scale, impact, directness, and timing. Traditional risk management approaches neglect to prepare for both compound disturbance that are not well defined or are novel and for non-stationary threats including future coastal scenarios impacted by sea level rise. Thus, a framework for evaluating the resilience of an infrastructure system rather than its’ risk tolerance is required.

In this work, resilience is defined as the ability of a system to prepare for, withstand, recover from, and adapt to future unknown disruptions. New York City NY, Gulfport MS, and Camp Lejeune NC were chosen as case study locations to evaluate a framework combining threat data and network analysis. For the current case studies threat data principally consists of flood depths from hydrologic models and the transportation network was used for analysis. Compound disturbances, consisting of GSSHA flood model output during Super Storm Sandy and randomized bridge failures were applied to the New York City roadway network. The impacts of this were analyzed using 10-minute travel time ego-nets around critical points in the city including hospitals, fire stations, and FEMA shelters. Compound disturbances near Camp Lejeune were represented on a regional level including NC-DOT regions one, two, and three with flooding data provided by FIMAN-T. For this location both the transportation network and the power grid were considered. In Gulfport, sea level rise was projected out to 2100 and used to drive an Adaptive Hydraulics 2D Shallow Water (AdH-SW2D) model of the area. This nonstationarity was added to with the inclusion of different river flow scenarios retrieved from StreamStats, a USGS tool. A similar effort was undertaken for Camp Lejeune, North Carolina.

This work aims to take knowledge and developed processes from the case studies described above to create a framework for resilience quantification. This framework will be able to take in threat data from a variety of sources and disciplines and apply this data any infrastructure network. The framework would enable forecasted or near-real time emergency response to natural hazards. Additionally planned infrastructure improvements and new construction could be efficiently evaluated for resilience during the design phase.