Abstract: Flood disasters triggered by extreme weather typically occur rapidly and extensively. Conventional monitoring methods struggle to satisfy the emergency monitoring of these disasters due to limitations such as small coverage, insufficient timeliness, and high operational risks. With the rapid advancement in domestic satellite technology, remote sensing technology offers advantages such as rapid response, wide observational range, and the objective representation of surface information. Accordingly, conducting real-time monitoring of extreme weather-triggered flood disasters, disaster assessment, and post-disaster reconstruction using multi-type and multi-frequency remote sensing data exhibits prominent advantages and high feasibility. In this study, a full-lifecycle emergency monitoring framework that integrates space-air-ground multimodal data was proposed, covering a whole service chain consisting of extreme event triggering, data programming and acquisition, intelligent image processing, disaster monitoring and assessment, and comprehensive decision-making sequentially. Based on high-frequency observations using domestic satellites and multispectral-synthetic aperture radar (SAR)-hyperspectral multimodal collaboration, combined with edge computing and deep learning, the proposed framework can rapidly extract the flooding range, identify infrastructure damage, and assess hazard risks. The framework was verified against the heavy rainstorm-triggered flood disaster occurring in Miyun District of Beijing City in 2025. The verification results indicate that the framework enables rapid emergency response, shortening emergency response time from days to hours and, thereby, providing support for government departments to make effective decisions.