| 14 Sep 2026
Global fire emissions impact on tropospheric chemistry and radiation in the Energy Exascale Earth System Model (E3SM)
Abstract. Wildfires release large amounts of trace gases and aerosols into the atmosphere, thereby playing a significant role in structuring tropospheric composition. Understanding and quantifying the interactions between fire, atmospheric chemistry, and climate has become increasingly important as wildfire activity has intensified in many regions over the past several decades. Here, we applied the Energy Exascale Earth System Model (E3SM) version 3, coupled with an interactive chemistry module as its default configuration and GFED5 fire emissions, to assess how contemporary wildfires influence atmospheric composition and radiation. We find that wildfires increase global annual average CO and O3 column concentrations by 16 ± 2 % and 6 ± 1 % (reported as mean ± standard deviation hereafter), respectively, during 1997–2022, with the largest increases occurring in wildfire-prone regions of the southern Amazon, central Africa, and tropical Asia. We also find that wildfires strongly promote interannual variations in CO and O3 at tropical sites, thereby improving consistency with in situ observations. Fire emissions increase global aerosol optical depth (AOD) by 7 ± 2 %, resulting in a reduction in net radiation at the surface by about 1.1 W m−2, primarily due to the direct attenuation of incoming solar radiation and indirect cloud effects associated with fire-emitted aerosols. These cooling effects are only partially offset by a small fire-induced ozone radiative warming at surface (+0.07 W m−2). Our analysis provides insight into the spatiotemporal influence of wildfires on global tropospheric trace gas and aerosol abundances and establishes several new diagnostics for evaluating coupled climate-wildfire models.
This manuscript presents a clear and generally well-structured assessment of the impacts of contemporary wildfires on atmospheric composition and radiation using E3SM version 3 with interactive chemistry. The analysis covers multiple atmospheric constituents and radiative effects over an extended period, and the results are generally interesting and relevant. I have a few minor comments and suggestions that I believe could further improve the manuscript.
Specific comments
Lines 86–90: In addition to black carbon (BC), brown carbon from wildfires deposited on snow and ice can also reduce surface albedo and affect the surface energy budget (e.g., Lin et al., 2014; Flanner et al., 2021)
Lin, G., Penner, J. E., Flanner, M. G., Sillman, S., Xu, L., and Zhou, C.: Radiative forcing of organic aerosol in the atmosphere and on snow: Effects of SOA and brown carbon, J. Geophys. Res.-Atmos., 119, 7453–7476, https://doi.org/10.1002/2013JD021186, 2014.
Flanner, M. G., Arnheim, J. B., Cook, J. M., Dang, C., He, C., Huang, X., Singh, D., Skiles, S. M., Whicker, C. A., and Zender, C. S.: SNICAR-ADv3: a community tool for modeling spectral snow albedo, Geosci. Model Dev., 14, 7673–7704, https://doi.org/10.5194/gmd-14-7673-2021, 2021.
Lines 118–123: How does the present study differ from these previous efforts of Earth system modeling studies, and what are its main novel contributions? A clearer explanation of the study's novelty would help establish its significance.
Lines 150–152: The statement that “the atmospheric chemistry and aerosol modules are implemented as separate components, with chemistry calculations performed independently of the aerosol module” is somewhat unclear. Does this mean that the two modules calculate their respective processes separately, potentially without fully accounting for the interactions between gas-phase chemistry and aerosols? Are any processes treated independently or calculated more than once? Since gas-phase chemistry and aerosol processes are physically and chemically coupled, please clarify how these components exchange information and represent their interactions in E3SM version 3. This clarification is important for understanding the model configuration and interpreting the results.
Model uncertainties: The manuscript would benefit from a more explicit discussion of uncertainties associated with the fire emission inventory and model parameterizations. For example, how sensitive might the results be to the assumed IVOC-to-POA mass ratio and uncertainties in the GFED5 fire emissions? Even a brief discussion of these uncertainties and their potential implications for the simulated atmospheric composition and radiative effects would strengthen the study.
Figure 8: Both simulations appear to overestimate O₃ concentrations, and the inclusion of fire emissions seems to exacerbate this bias. Could the authors discuss the potential reasons for the overestimation? In particular, it would be helpful to distinguish the contribution of fire emissions from potential biases in the model's background ozone chemistry, precursor emissions, and atmospheric transport. This discussion would also help clarify how the model can reproduce the observed interannual variability while retaining a systematic bias in mean O₃ concentrations.
Figure 9: The fire-induced change in surface O₃ appears to be smaller than the change in column-averaged O₃. What does this difference imply about the vertical distribution of the fire-induced ozone perturbation? Does it suggest that the ozone enhancement is concentrated in the middle or upper troposphere rather than near the surface? If possible, a vertical profile analysis or a brief discussion of the relevant mechanisms would help explain this result and its implications for surface air quality and radiative forcing.