PREDICTION OF BUILT-UP LAND CHANGES IN BANDAR LAMPUNG CITY

Abstract

Rapid population growth in Bandar Lampung City has driven the conversion of non-built-up land, highlighting the need to study future land changes to anticipate emerging challenges. This study aims to (1) map built-up land in Bandar Lampung City and analyze its changes from 2013 to 2023, (2) identify the most influential determinant variables affecting built-up land changes, and (3) predict the extent of built-up land up to 2045 through spatial modeling using multi-temporal imagery and physical-environmental and socio-demographic determinant data. A quantitative research method was employed, where built-up land was derived from Landsat 8 imagery using the Normalized Difference Built-up Index (NDBI) algorithm. The built-up land prediction model was developed using an integrated Markov Chain-Multilayer Perceptron-Cellular Automata (MC-MLP-CA) approach. The analysis revealed that built-up land in Bandar Lampung City increased from 9,410 ha (53%) in 2013 to 12,400 ha (70%) in 2023, marking an expansion of 2,990 ha (17%). The most influential variable in this change was proximity to roads. Model validation yielded an overall kappa value of 0.877, indicating high accuracy. Projections indicate that Bandar Lampung City's built-up land will expand to 16,010 ha (90.04%) by 2045. These findings highlight the importance of predictive spatial modeling in supporting proactive and sustainable urban planning strategies.