Abstract

Timely and accurate flood extent mapping is crucial forflood management and responses in order to mitigate flood consequences,especially in the context of climate change. This study applied NormalizedDifference Flood Index (NDFI) and Normalized Difference Flood in shortVegetation Index (NDFVI) to map the flood inundation rapidly using time-seriesSynthetic Aperture Radar (SAR) data, Sentinel-1. The flood maps were generatedbased on these two indexes calculated for pre and during flood images and thelow-vegetation inundated areas classified and added to the flood maps. TheSentinel-1 images are considered suitable and necessary as they are an opendata source and the high rate of revisit time (12 days). However, this approachcould be applied for other C-band SAR data as well. The flood season in 2018 inDong Thap province is chosen for the case study with 143 field survey pointsfor accuracy assessment. Our results prove that the applied methods were highlyaccurate and productive. As the Sentinel-1 data were collected for the entireVietnam and stored in the Vietnam Data Cube of Vietnam National Space Center,this method of rapid flood mapping is a potential and enable tool applying forthe entire Vietnam area.