Quantifying spatiotemporal dynamics of urban building and material metabolism by combining a random forest model and GIS-based material flow analysis

Understanding building metabolism is critical for guiding urban resource management and addressing challenges in urban sustainable development. Key attributes of buildings, including geolocation, footprint, height, and vintage, are crucial to characterizing spatiotemporal patterns of building metabolism. However, these attributes are usually challenging to obtain broadly and automatically, which obscures a comprehensive understanding and accurate assessment of urban metabolism. Moreover, the lack of a finer spatial pattern of these attributes shadows a spatially explicit characterization of material stock and flow in cities. In this study, we took Shenzhen—whose urbanization over the past three decades has been unprecedented in China and even around the world— has been taken as an example to develop a city-level building dataset based on a random-forest model and quantify the spatiotemporal patterns of material metabolism at relatively high spatial resolution (in 500 m × 500 m grids) by combing material flow analysis (MFA) with geographic information system (GIS). The results show that Shenzhen grew from a small town with 281.02 × 106 m3 of buildings in the 1990s to a mega-city with 3585.5 × 106 m3 of buildings in 2018 and expanded both outward and upward from downtown to suburban areas. The urban “weight” (material stock) increased from 92.69 Mt in the 1990s to 1667.8 Mt in 2018 and tended to be saturated, with an average growth rate of 9.5% per year. Spatially, the south-central areas were the largest container of material stocks and generated the most demolition waste. The spatially explicit maps of building three-dimensional (3-D) form and vintage provide detailed information for architectural conservation and could support the decision-making for urban renewal planning. The spatiotemporal patterns of in-use material stocks and potential generation of construction and demolition waste (CDW) provide a benchmark of environmental risk assessment and potential secondary resources to reduce “original” material consumption, which could help alter urban renewal to an environmental-friendly and sustainable trajectory.