Exploration of Vulnerability of Temperature Changes in Southeastern Coastal Islands of Bangladesh through the 2 Decades of Spatiotemporal Data

Bangladesh is one of the most vulnerable countries to climate change-related disasters and economic loss and damage. This study examines 20 years of satellite-derived land surface temperature (LST) data to investigate seasonal trends, changes in land use and land cover (LULC), and the relationship between temperature changes and the most common mangrove species in the Coastal islands of Bangladesh. The most noticeable temperature changes happened in the pre-monsoon and monsoon seasons. In December, on the other hand, there was a statistically significant cooling trend of −0.041 ℃ per year. At the same time, forest cover has been shrinking by an average of 26.36 km² per year, while coastal water bodies have been growing by 23.44 km² per year. Cluster analysis shows that temperatures change a lot from month to month outside of the pre-monsoon season. This suggests that the climate is unstable and could push the system beyond ecological thresholds. SARIMA modelling demonstrated 98.12% accuracy in predicting temperatures, highlighting the importance of temporal analysis in forecasting future stress thresholds. Species-specific temperature clustering shows how different mangrove species can handle heat: Ceriops decandra is more common in locations with higher temperatures, while Heritiera fomes is more common in areas with lower temperatures. These patterns show that ecosystem resilience is becoming less stable; therefore, we need to move from passive Conservation to proactive, species-informed, and thermally adaptive management practices.