Facility-scale greenhouse gas emission quantification at the Bremen steelworks using ground-based remote sensing

The Integrated Greenhouse Gas Monitoring System (ITMS) aims to establish an operational top-down monitoring framework for greenhouse gases (GHG) in Germany by combining atmospheric in situ and remote-sensing observations with atmospheric transport modelling and inverse estimation techniques. Power plants and large industrial facilities account for more than half of global anthropogenic CO₂ emissions and are therefore key targets. However, the limited temporal and spatial resolution of satellite GHG observations makes complementary ground-based measurements necessary for robust emission quantification at the facility scale.

This work contributes to ITMS by assessing the capability and uncertainties of quantifying GHG emissions from a major point source using ground-based observations and atmospheric transport modelling. The study focuses on the Bremen steelworks, comprising two blast furnaces and a blast-furnace-gas-fired power plant, emitting approximately 5 Mt CO₂ yr⁻¹ and accounting for nearly half of the city’s total emissions.

The campaign measurements conducted between April and June 2024 and 2025 targeted the plumes of the steelworks: Two portable Bruker EM27/SUN FTIR spectrometers measured column-averaged abundances of CO₂, CO and CH₄, while background values were provided by the Bruker 125HR FTIR spectrometer at the University of Bremen. Mobile zenith-sky DOAS observations of co-emitted NO₂ constrained plume width and trajectory, surface CO₂ concentrations were measured in situ, and wind profiles were obtained from a Doppler wind lidar. Plume transport was simulated with a Gaussian plume model and combined with excess CO and CO₂ measurements in an inversion framework to derive emission ratios and emission estimates.

The derived CO/CO₂ emission ratio is 3.46% ± 0.85%, consistent with emission inventories (3.33%, Umweltbundesamt). Constraining the model with real-time DOAS plume observations yielded preliminary emission estimates ranging from 40% to 179% of inventory values, with an average of 79% ± 49%. These results highlight both the promise and current limitations of ground-based remote sensing in reducing uncertainties of facility-level emission quantification.