Observational constraints reduce estimates of the global mean climate relevance of black carbon

Gunnar Myhre, B. H. Samset, Camilla W. Stjern, Øivind Hodnebrog, Ryan J. Kramer, Chris Smith, Piers Forster, Timothy Andrews, Oliviér Boucher, G. Faluvegi, Drew Shindell, Alf Kirkevåg, Trond Iversen, Dirk Olivié, Philip Stier, Duncan Watson‐ParrisResearch 2021 · DOI

Abstract

Black carbon (BC) aerosols have long been considered a significant driver of global warming, yet their precise impact remains uncertain due to incomplete knowledge of their sources, optical properties, and atmospheric behavior. This study constrains four global climate models with observational data on BC concentrations, vertical profiles, and absorption efficiency. We find that the global mean surface temperature change from anthropogenic BC is likely much lower than previously estimated, at approximately +0.03 ± 0.02 K. This reduction is primarily driven by atmospheric "rapid adjustments" and a correction of vertical distribution biases in models. Consequently, the 100-year Global Warming Potential ($GWP_{100}$) for BC is revised downward from 680 to 160 ± 120.

Introduction

Black carbon, or soot, is a potent light-absorbing aerosol produced by the incomplete combustion of fossil fuels and biomass. While its ability to warm the atmosphere is well-documented, quantifying its net effect on the global climate has been challenging. Previous estimates often relied on models that may have overestimated BC concentrations in the upper atmosphere and neglected the complex feedback loops known as rapid adjustments. This paper presents a refined assessment of BC's climate relevance by integrating aircraft observations and the latest emission inventories into climate simulations.

Addressing Model Biases: Vertical Distribution and Rapid Adjustments

Two critical factors have historically inflated the estimated warming potential of BC. First, many climate models exhibit a "vertical bias," placing too much BC in the upper troposphere where its radiative impact is most pronounced. Second, traditional metrics often focus on Instantaneous Radiative Forcing (IRF), ignoring rapid adjustments—short-term changes in clouds, humidity, and temperature profiles that occur before the ocean and surface have time to warm up.

To address these, the researchers utilized four models from the Precipitation Driver Response Model Intercomparison Project (PDRMIP): ECHAM-HAM-M7, GISS, NCAR-CESM1-CAM4, and NorESM1. They compared standard simulations (BC STD) using unconstrained profiles with vertically constrained simulations (BC VC) that match actual aircraft-measured profiles. These simulations were further scaled using the latest Community Emission Data Set (CEDS v2021) and observational estimates of the Mass Absorption Coefficient (MAC).

Results: A Significant Downward Revision

The study reveals that rapid adjustments, particularly changes in cloud cover, offset nearly 50% of the initial warming effect. When BC warms the surrounding air, it can "burn off" clouds or alter their distribution, leading to a cooling effect that counteracts the direct absorption of sunlight.

Figure 1: Comparison of radiative forcing components (IRF, Adj, ERF) and Global Warming Potential (GWP) between standard (STD) and vertically constrained (VC) models.

As illustrated in the figure above, the Effective Radiative Forcing (ERF)—which accounts for these rapid adjustments—is significantly lower than the IRF. By further applying vertical constraints to correct for upper-atmosphere overestimation, the ERF drops to 0.08 $W m^{-2}$, a reduction of over 50% compared to unconstrained simulations. This shift has profound implications for climate metrics; the $GWP_{100}$ for BC is now estimated at 160 ± 120, a sharp decline from the previous estimate of 680.

Implications for Climate Policy

These findings suggest that while reducing BC emissions remains crucial for public health and air quality, its role as a "quick fix" for global warming may be overstated. The net climate benefit of BC mitigation is smaller than previously thought, especially when considering that BC is often co-emitted with cooling aerosols like sulfates. The study highlights the need for climate policies to move beyond simple metrics and account for the complex atmospheric responses that determine the true impact of aerosol emissions.

Conclusion

By grounding climate models in observational data, this research provides a more realistic picture of black carbon's role in the Earth's energy balance. The significant impact of rapid adjustments and the correction of vertical biases necessitate a re-evaluation of how we prioritize short-lived climate pollutants in global mitigation strategies. Future work with larger model ensembles and a deeper focus on co-emission complexities will be essential to further refine these estimates.

Summary updated

Abstract

Abstract How emissions of black carbon (BC) aerosols affect the climate is still uncertain, due to incomplete knowledge of its sources, optical properties and atmospheric processes such as transport, removal and impact on clouds. Here we constrain simulations from four climate models with observations of atmospheric BC concentrations and absorption efficiency, and the most recent emission inventories, to show that the current global mean surface temperature change from anthropogenic BC emissions is likely to be weak at +0.03 ±0.02K. Atmospheric rapid adjustment processes are found to reduce the top of atmosphere radiative imbalance relative to instantaneous radiative forcing (direct aerosol effect) by almost 50% as a multi-model mean. Furthermore, constraining the models to reproduce observational estimates of the atmospheric vertical profile reduces BC effective radiative forcing to 0.08 W m-2, a value more than 50% lower than in unconstrained simulations. Our results imply a need to revisit commonly used climate metrics such as the global warming potential of BC. This value (for a 100-year time horizon) reduces from 680 when neglecting rapid adjustments and using an unconstrained BC profile to our best estimate of 160 ±120.
BibTeX
@inproceedings{myhre2021observational,
  author = {Gunnar Myhre and B. H. Samset and Camilla W. Stjern and Øivind Hodnebrog and Ryan J. Kramer and Chris Smith and Piers Forster and Timothy Andrews and Oliviér Boucher and G. Faluvegi and Drew Shindell and Alf Kirkevåg and Trond Iversen and Dirk Olivié and Philip Stier and Duncan Watson‐Parris},
  title = {{Observational constraints reduce estimates of the global mean climate relevance of black carbon}},
  booktitle = {Research},
  year = {2021},
  doi = {10.21203/rs.3.rs-691895/v1},
}
Figure