Incorporating a new biogenic flux estimate of carbon monoxide into the TM5 atmospheric chemistry model

Carbon monoxide (CO) plays an important role in tropospheric chemistry by reacting with hydroxyl radicals (OH) and thereby influencing the atmospheric oxidative capacity. While primary CO sources include fossil fuel combustion, biomass burning, and hydrocarbon oxidation, terrestrial ecosystems can also emit and consume CO through a range of biotic and abiotic processes. Traditionally, atmospheric models assume ecosystems as net CO sinks. However, this assumption is challenged by limited field measurements, outdated dry deposition schemes, and poor quantification of biogenic emissions, leading to uncertainties in the CO budget.  

We present the implementation of a new eddy covariance-based biogenic CO flux estimate in the TM5 atmospheric chemistry model. Two forward model simulations were performed for the period 2015–2020, using a traditional resistance-based dry deposition scheme and a new bottom-up estimate of biogenic CO fluxes derived from eddy covariance measurements. Simulated CO concentrations from both runs are evaluated against NOAA surface observations and TROPOMI satellite observations to assess whether the new biogenic CO flux representation improves TM5 model performance, particularly in capturing spatial and temporal variability and in representing spatial gradients in atmospheric CO.  

Preliminary results indicate an overall change of 195 Tg CO yr⁻¹ in prior global fluxes between the two forward runs. This change results from a reduced soil sink when the traditional dry deposition scheme is not applied, together with increased biogenic surface emissions in the eddy covariance–based prior estimate. The modified prior estimate increases surface CO concentrations over land by 7.0 ppb in the Northern Hemisphere (30°N–90°N) and 8.0 ppb in the tropics (30°S–30°N), while decreasing them by 1.1 ppb in the Southern Hemisphere (30°S–90°S). Further analysis is ongoing to quantify the potential improvement in model performance.