<![CDATA[Land cover changes over the past two decades have resulted in a significant decline in water infiltration capacity, particularly in regions experiencing high development pressure such as the Bolon Watershed in North Sumatra. This decline is closely associated with the conversion of natural vegetation to residential areas and built-up land. This study aims to predict water infiltration conditions in the Bolon Watershed over the next ten years using an integrated spatial-temporal approach. Three methodologies were employed: Cellular Automata-Markov Chain (CA-MC) for projecting land cover changes, Weighted Multi-Criteria Analysis (WMCA) based on Analytic Hierarchy Process (AHP) for multi-parameter assessment, and Artificial Neural Network (ANN) to capture nonlinear relationships among environmental variables. The results indicate that although areas with high infiltration capacity remain dominant, their extent tends to decrease. Conversely, areas with very low infiltration capacity show an increasing trend, reflecting the continuous pressure of urbanization. The most influential variables affecting land infiltration capacity are land cover, soil texture, slope gradient, and precipitation, with expert assessment results demonstrating acceptable consistency. This study emphasizes the importance of integrating predictive spatial methods in spatial planning to identify priority conservation areas. Through this approach, data-driven policy formulation becomes more accurate in preventing the degradation of ecological functions in water infiltration areas.]]>
Copyrights © 2026
