Identifying the effects of new technologies for extreme flood adaptation in China

Technological progress can help reduce the flood adaptation gap, but its sector-specific impacts remain unclear. We propose a text-mining-based structural equation modeling approach to analyze cross relationships between technology and perceived flood losses across multi-sectors, using 3.58 million flood-related posts during 1000-year-return extreme flooding in two Chinese cities. We assess the effects of four technologies: Unmanned Aerial Vehicles (UAVs), Life-saving Documents (LSD), AMap, and "Dragon Water" drainage (DW). Results show high heterogeneity in their effects—UAVs and LSD support multisectoral recovery, with UAVs driving other technologies, while AMap and DW target traffic and buildings. For casualties, communication, and water/electricity supply, UAVs (43.0%~62.2%) and LSD (19.5%~37.1%) rank higher in importance than rescue effort effectiveness (REE) (17.3%~28.5%). Technology adoption influences perceived losses through REE, with LSD having the strongest propagation path (coefficient: 0.577 ~ 0.579). This study offers a text-based framework for evaluating emerging technologies in flood adaptation and provide insights for the flood adaption planning of technology readiness considering different flood damage scenarios with resource constraints.