Spatial and Temporal Variability of Near-Surface CO2 and Influencing Factors in Urban Communities

Abstract

CO2 is the primary contributor to global warming, and also the most significant anthropogenic emission gas in cities. This study investigates near-surface CO2 spatiotemporal variability patterns at the community scale to address the critical gap in urban CO2 high-resolution measurement and promote urban carbon neutrality. Combining fixed and mobile monitoring across five representative communities (1-km2 coverage) with two-hour temporal precision and 20 m spatial resolution, results revealed average CO2 concentrations of 440–480 ppm, exhibiting bimodal diurnal cycles and highlighting spatiotemporal divergent emission behaviors. Three communities peaked during 17:00–19:00 LT, while two peaked during 08:00–10:00 LT. Spatial correlation analysis identified two dominant patterns: road-adjacent “externally dominated” hotspots and “internally dominated” zones with elevated intra-community levels. Spearman correlation analysis, Random Forest, and Geographically and Temporally Weighted Regression models quantified spatial morphology and element contributions, demonstrating that building morphology exerted time-varying impacts across communities. Meanwhile, external traffic contributed 18–39% to concentration variability, while internal traffic and energy consumption drove localized peaks. The findings indicated that apart from the emission sources, the micro-scale urban spatial design elements also regulate the near-surface CO2 distribution. This high-resolution approach provides actionable insights for optimizing community layouts and infrastructure to mitigate localized emissions, advancing carbon neutrality targeted urban planning.