Objectives  Flood disasters are often accompanied by cloud and rain. Time series data obtained before and after the disaster come from various sources, how to integrate multi-modal and multi-temporal remote sensing images to monitoring flood is the general trend.

  Methods  We propose a flood monitoring method using multi-modal sequence remote sensing image integration registration and automatic change detection. The depth features and semantic information can be used to realize automatic and high-precision registration of optical and SAR(synthetic aperture radar) images before and after disasters. Based on geographic information and multi-time series remote sensing images, we realize flood change monitoring and submergence information extraction.

  Results  The method has been effectively verified by flood in Anhui, China in July 2020 and in Henan, China in July 2021, and also can achieve faster acquisition of post-disaster damage information.

  Conclusions  This work solves the problem of emergency flood monitoring in the case of rainy weather, and puts forward several suggestions on the application of remote sen-sing technology to disaster events in our country.