Quantifying CO2 emissions from an industrial complex using eddy covariance flux measurements

Industrial complexes are significant sources of anthropogenic CO₂ emissions, yet global emission inventories often struggle to capture the fine-scale spatiotemporal variability. This study quantifies CO₂ fluxes in the Sihwa National Industrial Complex (SIC), a major manufacturing cluster in southwestern Seoul, South Korea, using one year of continuous eddy covariance (EC) observations (Feb 2024 Jan 2025). A fixed EC tower (14 m height) equipped with an open-path CO₂/H₂O analyzer and a 3D sonic anemometer provided high-frequency flux data, processed using IQR filtering, U* thresholding, and multi-scale gap-filling. Flux footprint modeling (FFP) was used to separate contributions from industrial and green areas. Our results show that industrial zones acted as persistent net CO₂ sources (mean: 228.3 tC km^-2 month -1 ; max: 376.8 tC km^-2 month^-1 ), exhibiting strong diel and weekly cycles linked to operational and energy-use patterns. In contrast, adjacent green spaces functioned as seasonal CO₂ sinks (min: -136.7 tC km^-2 month ^-1 ), particularly during the growing season. Compared to the ODIAC top-down inventory, EC-based emissions were lower by up to 60% across all months. This discrepancy reveals a fundamental limitation of inventory-based approaches: their reliance on static emission factors, coarse-resolution proxies (e.g., nightlight), and the absence of real-time, local activity data, which often leads to significant overestimation in complex environments like industrial zones. This study underscores the necessity of integrating ground-based flux observations into national and global emission accounting frameworks to improve the accuracy, transparency, and policy relevance of industrial CO₂ inventories.