Work

From traffic-flow capacity to recovery capacity: An episode-based framework for multi-sector urban resilience

Year: 2026

Type: article

Abstract: Urban systems experience disruptions that differ in magnitude, duration, and final recovery outcome. Existing resilience measures commonly emphasize performance loss, recovery duration, or trajectory area, but they do not directly represent the peak rate at which a disrupted system recovers. Inspired by the traffic-flow view of capacity as a dynamic, state-dependent peak rate, this paper develops an episode-based recovery-capacity framework. A Double Quadratic Queue (DQQ) procedure is used as an episode filter to identify critical transition points, and a Linear-Exponential Adaptation (LEA) model represents the recovery trajectory. Single-period recovery capacity is defined as the maximum healing rate, R c a p = a b e − c , while multi-episode stress–strain relationships provide complementary descriptive measures of disruption sensitivity and transition loss or gain. Model comparison using Google COVID-19 Community Mobility Reports shows that LEA generally provides lower fitting and prediction errors than benchmark recovery models. Extreme-heat episodes in Phoenix and post-earthquake infrastructure restoration in Japan are included as transfer illustrations with case-specific variables and interpretations. The resulting framework is descriptive and diagnostic rather than causal, and it provides a focused connection between observed recovery episodes and dynamic capacity analysis.

Source: Transportation Research Part C Emerging Technologies

DOI: 10.1016/j.trc.2026.106063