Assessing Tropical Cyclone Destruction Using Landsat Satellite Imagery: A Case Study in Bangladesh

Tropical cyclones are extreme natural catastrophes and are anticipated to intensify strength and recurrence under upcoming climate change consequences in our coastal regions through Bangladesh. This study has tested and developed methods for evaluating considerable damages affected in 2020 by cyclone Amphan in Satkhira district (3817.29 km2) in Bangladesh from 30m spatial-resolution satellite imageries. Iso-cluster unsupervised classification procedures are applied to prepare maps in pre- and post-cyclone Landsat 8 imagery. A difference between the pre- and post-cyclone image classification and NDVI analysis give us evidence about the significant landscape changes at the Cyclone Amphan impact on the study area. The total accuracies for pre- and post-cyclone images classification were 70% and 68% respectively. The outcomes indicate that 35.09% of the study region was remarkably influenced by cyclone Amphan. Statistical analysis indicates that water bodies have dramatically increased 17.65% in different land cover areas due to the strong water surge. On the contrary, the classified image showed that the vegetation and agricultural land decreased 4.13% and 10.82%. Thus, land covers have inundated due to heavy flash during the cyclone. Barren land also decreased 1.80%, and low land areas were also flooded 0.36% due to heavy water flashes. NDVI analysis indicated no destruction of any vegetation species, but vegetation damage is considerable. The developed method may be imposed in future to measure the impacts produced by cyclones in Bangladesh and to make policies for the commitments of cyclone disaster controlling and management.