Atmospheric hydrogen budget: an evaluation using chemistry-transport models and a box model

Hydrogen fuel, a green transition option and a cleaner alternative to fossil fuels, has an indirect greenhouse impact through atmospheric reactions of “leaked” hydrogen. Sand et al., 2023 used six different chemistry-transport models (CTM) to estimate a Global Warming Potential over a 100-year time horizon (GWP-100) for hydrogen of 11.6 ± 2.8, in range with similar studies. In this study, we extend those analyses by investigating the atmospheric production and loss terms of hydrogen in the CTMs. Specifically, we compare formaldehyde (HCHO) and the hydroxyl radical (OH) concentrations. Then we develop a box model that can be used for quickly evaluating the impact of the different sources and sinks on atmospheric concentration and isotopic composition of H₂ from a global perspective.

Atmospheric production of hydrogen through photo-oxidation of methane and volatile organic compounds represents roughly 60% of the total production. To compare the atmospheric production in the models, we evaluate HCHO (produced during photo-oxidation). A preliminary comparison between the global mean model-derived tropospheric HCHO and TROPOMI-derived HCHO suggests that all models other than WACCM perform reasonably well. Generally, models tend to overestimate HCHO values over land and underestimate HCHO concentrations over the oceans. WACCM has very low HCHO values compared to TROPOMI and the other models.

The two primary removal mechanisms are soil uptake (65-85%), and atmospheric oxidation by hydroxyl radical (OH). Among the models, OsloCTM3 and WACCM have higher OH concentrations compared to GFDL, INCA and UKCA. Direct measurements of atmospheric OH concentrations are lacking due to the short lifetime of the OH radical. Therefore, we used CO and NO₂ concentrations as a proxy to evaluate the models. Compared to satellite values (TROPOMI for NO2 and MOPPIT for CO), models seem to generally overestimate NO₂ and underestimate CO. These results are discussed within the context of the OH radical and atmospheric lifetime of H₂.

Then we present a simple box model that is developed using CTM results for studying the atmospheric budget of H₂. Reconstructions of hydrogen concentrations using ice-core records from the South Pole over the last 150 years show an increase in H₂ concentration of ~200ppb, likely due to increased methane oxidation and anthropogenic emissions. We use time-varying emissions in our box model to replicate this time evolution since the pre-industrial period.

The box model also contains a framework for studying hydrogen isotopic composition. Each of the sources and removal processes of H₂ have distinct isotopic signatures. This allows for the evaluation of concurrent changes in atmospheric concentrations and hydrogen isotopic compositions for each source/sink contribution, leading to a more robust evaluation of the hydrogen budget in the atmosphere.