Near-Real-Time Emission Characterization for Major Industrial Sectors Using Multisatellite and Base-Year Emission Inventories

Near-real-time (NRT) characterization of industrial emissions is crucial for tracking dynamic emission patterns and informing timely regulatory responses. However, existing methods rely heavily on continuous emission monitoring systems (CEMS), which are often limited in data availability, coverage and data quality. This study introduces a novel approach that integrates multisatellite fire radiative power (FRP) observations with a base-year emission inventory (BY-EI) to improve source identification and emission estimation. Compared to previous methods, our multisatellite strategy increased the number of identified industrial point sources by 38%. Fire radiative energy (FRE)-emission response models were developed by linking satellite-derived FRE with emission data from identified sources. These models enabled daily emission estimates with strong performance, especially for SO2 (R2 = 0.895) and NOx (R2 = 0.855). For sources without direct FRP detection, sectoral FRE-based temporal profiles captured emission variability effectively (Pearson’s r > 0.6, mean square errors (MSE) at 10–5 level). This approach successfully identified high-emission industrial sectors, including cement, ceramics, and steel industries, expanding the scope of detectable sources while reducing reliance on CEMS data. The findings provide a new framework for NRT industrial emission characterization, supporting the timely updating of industrial emissions and refined pollution control strategies and decision-making.