Data-driven quantification of the resilience of enforcement policies under emergency conditions: A comparative study of two major winter storms in Buffalo, New York

The disruptive impacts of climate change are manifesting themselves in the form of unexpected and high impact extreme weather events. Their negative impacts on the residents of affected regions can be partially mitigated by government authorities through the deployment of effective real-time disaster response strategies. This study aims to investigate the time-dependent change in the regulatory power of transportation related policies during disruptive and deadly winter storms. A comparative study using the disruptive events of the November and December 2022 blizzards in Buffalo is conducted,aiming to identify residents’ trip making decisions during enforcement and recovery phases, using different time scales and performance of various transportation system indicators. We introduce a new metric namely, the Loss of Resilience of Policy (LoRp), drawing inspiration from the Loss of Resilience framework used in evacuation studies. This new performance measure is conceptualized to measure the time-dependent potency of enforcement policies from their initiation to their cessation. Results based on these new performance indicators are analyzed using a change point detection and spatial modeling framework on a neighborhood level to associate the calculated LoRp values with spatial and socioeconomic variables. The results are used to understand the effect of these co-variates on LoRp, and study the main factors affecting resilience of various policies. Furthermore, this study introduces a regression model linked to storm weather data to predict the rate of change and the amplitude of LoRp. This regression model serves as a tool for policymakers to preemptively adjust enforcement policies in a timely manner with the goal of improving responses to extreme weather events and reducing unnecessary delays.