View article

The Sundarbans mangrove forest is under constant pressure due to both natural and anthropogenic stressors. It is vulnerable to periodic cyclonic disturbance that cause short and long-term adverse impacts. This study assesses the changes in land cover caused by cyclone Roanu that struck the Sundarbans on 21 May 2016. Landsat imagery and two supervised classification algorithms, maximum likelihood (ML) and spectral angel mapping (SAM), were been used to analyze changes in land cover. In addition, these algorithms were compared in terms of classification accuracy. The results show a considerable change in land cover as a result of cyclone Roanu. Moreover, consistency between both classification methods has also been observed in spatial changes particularly for increase of water areas. Hence, it can be stated that the SAM was more accurate in detecting land cover changes of Sundarbans. In addition, the finding of this study has demonstrated the devastation of cyclone on Sundarbans. Therefore, these results may be helpful to maintain the quality of coastal ecosystem, protecting coastal lives, property, and livelihoods for properly managing the world heritage Sundarbans and its surrounding areas.