Understanding residents' responses towards disaster-related urban emergency management: A case study of typhoons in Shenzhen, China

Abstract

A comprehensive understanding of residents' responses to disaster emergencies is crucial for developing effective emergency strategies, whilst this critical problem remains underexplored in the existing literature. This study proposes an analytical framework that leverages publicly available social media data to model residents' behaviours in response to disaster emergencies and capture their underlying determinants. The framework integrates a hazard-based model and a binomial logistic regression model to examine how residents' response time and emotional states are associated with emergency measures, hazard characteristics and demographic factors. The framework is applied to examine public responses to five typhoons that affected Shenzhen, China, from 2018 to 2023. Our findings reveal notable heterogeneity in response patterns across different typhoon events, with behaviours primarily determined by hazard characteristics and emergency measures, rather than gender. Specifically, residents' responses are closely linked to typhoon-induced meteorological characteristics, underscoring the need to tailor emergency strategies to these features rather than relying solely on typhoon intensity metrics. In addition, emergency measures, including early disaster warnings, work suspension and increasing emergency funding, are found to improve response efficiency and reduce negative emotional reactions. The empirical insights from this study inform the development of context-specific, adaptive emergency strategies for Shenzhen and similar regions, while the proposed framework offers a scalable approach for analysing public responses more broadly.