Identification of land-cover features's phenology using multitemporal Sentinel-1 data: A case study in Hanoi, Vietnam

Hanoi regions are experiencing rapid land cover changes caused by human-included land degradation and extreme climates events. Identification the changes in vegetation phenology and in land-cover play an integral part for management and urban planning. Multi-temporal satellite image data are accepted as being the best material to map these changes, especially SAR images. SAR images are unaffected by weather conditions, clouds and operating all day and night. Currently, the Sentinel-1 data of European Space Agency (ESA) are free. The Sentinel-1A satellite carries a C-band Synthetic Aperture Radar, spatial resolution is 10 meters and acquisition cycle is 12 days. In the article, the authors present the correlation between the change of rice phenology and backscatter value on multi-temporal Sentinel-1 images. At the same time, the authors also propose a method of determining land-cover objects of Hanoi Vietnam on multi-temporal SAR images base on integrating changed/unchanged objects, mean backscattering value and rice phenology. The accuracy achieved 89.6%.