Predicting land use/land cover changes using CA-ANN model in the Gia Lam District, Hanoi City

Gia Lam is a district in the eastern part of Hanoi, Vietnam, undergoing rapid development and urbanization. Predicting land use/land cover changes is crucial for supporting policymakers in designing effective local development strategies. This study aims to forecast land use/land cover changes in Gia Lam District through 2028. Sentinel-2 optical satellite imagery serves as the primary data source, with the classification of five land use/land cover types at different times performed using the Random Forest (RF) algorithm, and forecast results generated using a Cellular Automata (CA) model combined with an Artificial Neural Network (ANN) model. The results indicate a significant expansion of built-up areas, projected to cover 68.92% of the natural area by 2028. On the contrary, the area under annual vegetation is expected to decline to just 7.71%. These findings provide valuable insights for researchers and land use planners.