ID: 40
Presenting Author: Minhee Je
Status: pending
This study develops a spatially explicit carbon emission prediction model for Korean cities using GIS and machine learning, serving as a decision-support tool for urban Environmental Impact Assessment
Cities are critical sources of global carbon emissions in achieving carbon neutrality goals. However, urban carbon emissions are spatially heterogeneous and vary significantly by urban typology. This study develops a spatially explicit carbon emission prediction model for major Korean cities with diverse geographical and functional characteristics.
Targeting metropolitan cities with populations exceeding one million in South Korea, the study integrates GIS-based spatial analysis with machine learning algorithms to predict building sector carbon emissions. Urban spatial variables including land use, building density, and population serve as predictors. Multiple machine learning models are compared to identify the most effective method for analyzing spatial emission patterns and determinants.
Findings quantify emission characteristics by urban typology, providing empirical evidence for spatially differentiated mitigation strategies. The model serves as a spatial decision-support tool applicable to urban Environmental Impact Assessments and climate impact assessments, enabling policymakers to identify emission hotspots and prioritize interventions. This approach strengthens EIA frameworks by integrating spatially explicit, data-driven analysis, supporting more effective and equitable urban climate action planning.
Minhee Je is a Researcher at the Building Research Division, Korea Institute of Civil Engineering and Building Technology (KICT), specializing in urban environment spatial analysis.
Coauthor 1: Seunghyun Jung