Building-level demolition and material output forecasting using 4d-GIS: a case study of Kitakyushu City, Japan

Abstract This study proposes a disaggregated modeling approach to estimate building-level demolition probabilities and project future material output. Many existing Material Stock and Flow Analysis (MSFA) studies rely on aggregated spatial units at national, prefectural, or municipal levels, which limits their ability to capture intra-urban variation. To address this issue, a Weibull Accelerated Failure Time (AFT) model was applied to building-level data in Kitakyushu City, Japan, incorporating structural, locational, and demographic variables. Demolition status was identified using geospatial overlays of building data from 2010 to 2018 and linked with spatial attributes such as land use zones, slope angle, and aging rates. The model was used to estimate demolition probabilities and simulate material output through 2040. Results indicate that buildings in aging and less accessible areas are more likely to remain despite the population decline, suggesting a growing risk of vacancy. The framework provides flexible spatial aggregation beyond administrative boundaries and can be introduced incrementally, making it suitable for areas with limited data availability. The study also highlights the importance of including year-of-construction data in building inventories. It will support sustainable urban development and enhance the framework, allowing for more strategic planning in the context of urban shrinkage and circular resource management.