BADAG(Building-stock Advanced Dynamic Applying Geospatial) Framework : High-Resolution Gridded Estimation of Future Building Stock

The building sector is essential for integrated global climate action, requiring a balanced approach that simultaneously addresses adaptation to climate risks and mitigation of greenhouse gas emissions. From an adaptation perspective, securing sufficient energy for cooling and heating is critical to reduce temperature-induced climate risks under extreme heat and cold conditions. From a mitigation perspective, substantial reductions are necessary not only in operational energy consumption with low demand strategies but also in the embodied carbon associated with retrofitting existing buildings and constructing new infrastructure. To support these dual climate targets, most integrated assessment studies initiate the projection of future energy and material demands by estimating building floor area, which serves as the fundamental proxy for quantifying service demand and material intensity.

However, existing studies predominantly relying on national-level variables are overly simplistic, as they typically model floor area solely as a function of income and population. This approach fails to capture the spatial heterogeneity within countries. In particular, it neglects the dynamic changes in floor area driven by increasing population density during urban growth. As a result, these models cannot capture how distinct urban forms interact with local climates to drive energy demand, limiting the feasibility of spatially explicit climate strategies

To address these limitations, this study proposes the BADAG (Building-stock Advanced Dynamic Applying Geospatial) framework, a bottom-up approach for estimating future building stock at a 1 km resolution under SSP scenarios. We constructed a comprehensive global spatial database integrating gridded socioeconomic indicators with building attributes from the Global Human Settlement Layer (GHS-OBAT). Our grid-level estimation model analyzes key determinants of floor area demand, simulating the non-linear dynamics linking floor area intensity to changes in population density and household size. Additionally, by leveraging regional correlations between floor area density and urban morphology defined by Local Climate Zone (LCZ) categories, we projected future urban structures. A rigorous calibration process was also implemented to correct potential underestimations in satellite-based datasets.

Applying this framework reveals significant divergences from conventional projections. In the Global South, our model estimates a lower total floor area than previously projected, suggesting that traditional methods overestimated stock by neglecting the limiting effects of increasing population density on per capita space. Conversely, in the Global North, total floor area is projected to increase despite slower growth, driven by shrinking household sizes and lower urban densities. Consequently, these structural shifts lead to a relative increase in cooling and heating energy demand in the Global North and a decrease in the Global South compared to conventional baselines.

These findings suggest that previous assessments may have misallocated climate risks and mitigation burdens due to inaccurate demand baselines. By providing a refined, spatially explicit estimation of building stock, this study demonstrates that advancing floor area projections is a fundamental prerequisite for valid integrated assessment. This enhanced projection enables stakeholders to correctly identify interdependencies between mitigation (operational and embodied emissions) and adaptation (energy requirements), ensuring strategies are based on realistic future urban contexts under SSP scenarios.