Data-driven insights into flood disasters: Evaluating the impact on residents' emotions and living spaces in the Beijing-Tianjin-Hebei region

Abstract

Frequent urban floods pose growing threats to residents' daily lives, highlighting the need to better understand their impacts on human well-being and everyday living environments. Previous studies have primarily focused on identifying flood-affected areas and analyzing residents' emotional responses. However, they often overlook how disaster severity relates to emotional reactions and the varied flood responses of different living spaces—leaving the differentiated impacts on residents' daily lives insufficiently understood. To address this gap, the study uses nighttime light data to quantify flood severity and relates it to residents’ emotional responses captured from Weibo posts. It further evaluates the flood performance of different living spaces—identified via POI and street view images—revealing spatial heterogeneity across urban areas. The results show that residents in severely affected areas exhibited more negative emotions, and this tendency became even stronger during the later stages of the disaster. Differences in infrastructure and mobility demands are linked to quicker recovery in residential and industrial zones, while commercial and tourist areas recover more slowly. Urban villages are heavily impacted, and those with poorer building quality in suburban areas exhibit lower flood resilience. Overall, this study integrates multi-source data to analyze how flood severity relates to residents' emotional responses and the varying resilience of different living spaces, offering deeper insight into the impacts of urban flooding on residents' daily lives.