Flood Monitoring Using Multi-temporal Synthetic Aperture Radar Images

摘要

The rise of water level in the basins of rivers is the most common cause of floods in the world. This can be occurs due to heavy rains, spring snowmelt, wind surge, destruction of dams, rubbish, straightening of rivers or deforestation. Satellite observing systems are the main tool for solving problems such as monitoring of flooded areas. Since it is often not possible to use optical images due to high cloud cover, data from radar satellite imagery is used to ensure all-weather satellite monitoring of the dynamics and effects of flooding. The paper proposes a computer information technology for satellite monitoring of floods based on multi-temporal synthetic aperture radar (SAR) data. The technology allows you to determine flood areas and quickly get floodplain area. In this paper we consider the possibility of remote detection of flooding in the Transcarpathian region in December 2017. Four multi-temporal radar images from the satellite Sentinel-1 were used for flood detection. Dates of shooting are 14, 17, 18 and 20 December 2017. The first image was taken before the flood, the next two - in the midst of a natural disaster, the last - after the decline of the water level. All presented radar images have the dual polarization VV+VH. The studies have shown that satellite radar images provide an opportunity to qualitatively determine the flooded area in the Transcarpathian region of Ukraine. The proposed methodology made it possible to qualitatively determine more than 2000 ha and identify damage areas after the flood. Using archival data in the methodology allows you to build cartograms for the previous period and identify regularity of flood areas.