Critical facility accessibility rapid failure early-warning detection and redundancy mapping in urban flooding

System-wide functionality failure emerges when a collection of components collapse simultaneously. Aiming to detect the early warning of the rapid critical facilities accessibility loss after flooding, this paper presents a network-theory-based detection model to identify the critical point after which the rate of road access loss to critical facilities starts to accelerate. The road network of Harris County, Texas during Hurricane Harvey 2017 flooding and simulated 500-year flooding (1,000 flooding sequences) is used for the case study. We also considered the travel radius for reaching critical facilities (e.g., 5 miles travel radius for hospitals, and 4 miles for groceries and pharmacies) in the accessibility assessment to show the flood impact on daily life. Additionally, we link the critical threshold to the spatial network to identify the corresponding critical roads whose disruption will likely trigger the rapid critical facility accessibility loss. This result will facilitate the decision-making on targeted infrastructure protection. Moreover, we map the number of critical facilities that each network component has access to within reach and reveal the resource redundancy inequity and risk disparity across the network. Integrating with the social vulnerability mapping, our result can help inform the resilience planning toward an equitable community.