Rapid community response and resilience to typhoon Doksuri in coastal Fujian: insights from large-scale human mobility data

Understanding human adaptive behavior and community resilience during typhoons is essential for developing human-centered disaster mitigation strategies. However, research on cross-city community responses and dynamic resilience during typhoons remains limited.

This study examines communities’ mobility responses and resilience patterns during Typhoon Doksuri in 2023 using anonymized mobile data from 3.7 million users across six coastal cities of Fujian, China. The study developed a novel Disturbance-Response-Recovery (DRR) framework for assessing dynamic resilience that integrates disaster hazard factors with mobility patterns during typhoons, offering more specific and empirical grounded insights. The results demonstrate rapid pre-disaster response and post disaster recovery in Fujian: 70% of communities responded within two days before landfall, community mobility intensity dropped by 80% on the landfall day, and 90% of communities recovered within three days post-impact. Pre-disaster community response variations were significantly associated with early warnings, typhoon track proximity, localized rainfall and wind intensity and historical typhoon experience, while recovery dynamic disparities were significantly predicted by rainfall, economic capacity (GDP) and road density. 

The results further revealed a significant spatial clustering pattern of community resilience that correlated with geographical location, disaster progression, and economic levels. Inland communities’ heightened sensitivity to disturbances, urban communities and towns that endured heavy rainfall demonstrated enhanced recovery capacity, and communities with lower economic status showed both higher sensitivity and lower recovery capacity. These findings provide valuable references for improving community resilience and optimizing disaster management strategies.