Change Detection Based Flood Mapping of 2015 Flood Event of Chennai City Using Sentinel-1 SAR Images

Chennai, the capital city of Tamil Nadu state, India experienced a major disastrous flood during Nov-Dec 2015. The city is characterized by mixed land use with high built-up density. The freely available C-band Sentinel-1 temporal GRDH SAR images are used to analyze this flood event. We used the co-polarized (VV) SAR images for mapping the flood area. A change detection technique (Normalized Change Index (NCI)) which uses a pre-flood image having the same image characteristics as the flood/post-flood images is implemented on the temporal SAR images. An automatic flood mapping approach, i.e. NCI combined with thresholding is used to identify the flood area. This approach has underestimated the flood area for this particular study area which is due to the use of moderate resolution SAR images for high-density urban flood mapping. Despite underestimation, Sentinel-1 SAR images are very useful to get the first insight of flood condition in the high density urban areas.