Extending the theory of planned behavior to explain Intra City migration during extreme rainfall disasters: Evidence from Zhengzhou, China

With climate change intensifying extreme rainfall events, understanding the mechanisms shaping migration motivation during disasters is critical for improving urban resilience governance. Using Zhengzhou's catastrophic“7·20” rainstorm as a case study, this study develops an Extended Theory of Planned Behavior (ETPB) to systematically analyze how five key variables—attitude, subjective norms, perceived behavioral control (PBC), risk perception, and social networks—affect migration motivation. An online questionnaire yielded 716 valid responses, capturing socio-economic attributes, migration experiences, and resettlement choices, as well as measures of these five constructs. Descriptive results show that about 30 % of respondents engaged in short-term migration, with nearly half of these movements being non-government-led, highlighting the importance of spontaneous and socially driven relocation. Women and older adults not only had lower migration rates but also significantly weaker migration motivation and lower PBC scores, indicating a heightened risk of involuntary immobility. Building on these insights, the ETPB model was constructed and validated using structural equation modeling. Results reveal that attitude and subjective norms significantly and positively influence migration motivation, while PBC is not significant. Incorporating risk perception and social networks increased the model's explanatory power to 64.7 %, with both exerting independent, significant positive effects. This study extends TPB to extreme rainfall contexts, identifies the multidimensional drivers of migration motivation, and offers actionable guidance for integrating community mobilization, risk communication, and gender- and age-sensitive measures into urban emergency planning and resilience strategies. • Collected migration behavior data from 716 Zhengzhou residents affected by the 2021 “7·20” urban flood. • Nearly half of migration was non-government-led, driven by social ties. • Identifies women and older adults as vulnerable to involuntary immobility. • Extends TPB by adding risk perception and social networks to model disaster migration.