Quantifying Wetland Loss in Bhopal Amidst Urban Sprawl: A Machine Learning Approach Using Multispectral Imagery

The process of urbanization has had a significant impact on the distribution of wetlands. This research aims to shed light on the interconnected dynamics between urban development and wetland alterations, providing valuable insights into environmental sustainability. The study will specifically focus on evaluating the performance of Machine Learning Algorithms in accurately mapping LULC (Land Use and Land Cover) from satellite datasets, i.e., optical (Sentinel-2), Landsat 8. By doing so, a deeper understanding of the complex interplay between urbanization and wetland distribution can be achieved. To achieve this, Machine Learning (ML) Algorithms like Random Forest (RF), Support Vector Machine (SVM), and Minimum Distance (MD) are applied and tested. This allows for an in-depth analysis of the spatial aspects of urbanization and its subsequent impact on the distribution of wetlands in the specific context of Bhopal. Despite the presence of multiple wetland types in the region, the study acknowledges the challenges in classifying them. We found a 51.29% decrease in wetland area and a 29.57% increase in urban area from 2016-2021. The results also indicate that the Sentinel-2 dataset outperformed the Landsat dataset, emphasizing the significance of visual information. In essence, this research contributes valuable insights into the complex dynamics of urbanization and wetland alterations, addressing the complexities involved in classifying diverse wetland types in the Bhopal region.