Abstract

In Bangladesh, the mangrove ecosystem of the Sundarbans region is vital to both the environment and the economy. Natural disasters, especially cyclones, cause significant damage to mangrove vegetation. To assess the impact of the cyclone on the Sundarbans, a long-term dataset focused on cyclone is required. Mapping mangrove forests using satellite imagery is an effective technique to find crucial information for mangrove forest conservation and management. We used Google Earth Engine (GEE) and Landsat 5, 7, and 8 satellites to assess the mangrove vegetation changes due to the cyclone during the study period. In this study, we analyzed the two indexes, Enhanced Vegetation Index (EVI) and Normalized Difference Vegetation Index (NDVI) for immediate pre-cyclone and immediate post-cyclone time for 8 cyclones that happened over the Sundarbans mangrove forest between 1990- 2020. From the statistical analysis, we found that cyclone wind speed is significantly negatively correlated with a lower NDVI and EVI value at a 95% significant level (0.05 confidence level). Among 8 cyclones the most catastrophic damage happened during cyclone SIDR (2007), on this time NDVI value decreased from -0.41223 to -1.0, and wind speed was the highest among the 8 cyclones, which indicates the devastating situation in the Sundarbans.

Findings:

• This study assessed that Bangladesh is facing cyclones almost once a year, but the Sundarbans have been facing cyclones frequently for a decade. Among them, only the strong cyclone significantly destroys the Sundarbans. Though these mangrove forests face devasting situations lots of times but these can re-generate their ecosystem. 

• The vegetation coverage of Sundarbans mangrove is so dense, so the vegetation index works significantly well. During the study period (1990-2020) the vegetation index was seen dynamically for every single cyclone. That means Sundarbans does need any man-made support to rebuild their eco-function. 

• During this study, it was found that mangrove forest vegetation analysis for natural disasters has been significantly hampered by tidal influences and cloud coverage. Using high-resolution radar active remote sensing data may solve these issues. 

• Based on 30 years of study, it revealed that the Sundarban's vegetation has statistically correlated with wind speed. Due to extreme speed during SIDR the change was noticeable among 8 cyclones. 

Abstract Goes to Here:

Abstract Goes to Here: