
==== Front
Proc Natl Acad Sci U S A
Proc Natl Acad Sci U S A
PNAS
Proceedings of the National Academy of Sciences of the United States of America
0027-8424
1091-6490
National Academy of Sciences

38277442
202322594
10.1073/pnas.2322594121
commCommentaryearth-sciEarth, Atmospheric, and Planetary Sciences413
437
Commentary
Physical Sciences
Earth, Atmospheric, and Planetary Sciences
Aerosols hold the key to recent and future Pacific warming patterns
McMonigal Kay ktmcmonigal@alaska.edu
a 1 https://orcid.org/0000-0001-5722-7036

aCollege of Fisheries and Ocean Sciences, University of Alaska Fairbanks, Fairbanks, AK 99775
1Email: ktmcmonigal@alaska.edu.
26 1 2024
6 2 2024
26 7 2024
121 6 e2322594121Copyright © 2024 the Author(s). Published by PNAS.
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).
==== Body
pmcOver the past 40 y—a period when most of the globe has warmed—the eastern tropical Pacific sea surface temperature (SST) has mysteriously cooled. Climate models forced by realistic anthropogenic greenhouse gas and aerosol forcings generally do not reproduce the observed eastern tropical Pacific cooling. For any single model run, this is unsurprising because random internal variability can lead to different patterns of warming (1). However, even considering climate model large ensembles, very few model runs can reproduce the observed tropical Pacific SST pattern (2, 3), suggesting that the models may be biased. Understanding why many climate models fail to simulate the eastern tropical Pacific cooling, and what it implies for future SST trends, is key to improving climate projections. A new study in PNAS (4) shows that an increase followed by a decrease in anthropogenic sulfate aerosol emissions can produce a multidecadal cooling pattern in the eastern tropical Pacific. This response hinges on both a fast and a slow response, which interact to yield a surface and subsurface cooling. Model biases in the location of the intertropical convergence zone, the strength of stratus low cloud-SST feedbacks, or ocean stratification may lead to an overall biased response, possibly explaining the inability of many climate models to simulate the observed Pacific SST trend. This detailed, process-based study opens a door for better understanding and projecting the future state of the equatorial Pacific.

Global sulfate aerosol emissions increased from 1920 to 1980, and have decreased since, largely due to air quality regulations in North America and Europe (5, 6). Anthropogenic aerosols alter climate by cooling the surface through direct aerosol-radiation interaction (7, 8) and recent work has shown that anthropogenic aerosols can alter SST trend patterns in the tropical Pacific (9). However, the timescales and mechanisms of the tropical Pacific response to varying sulfate aerosol emissions were not well understood. Hwang et al. investigate the effects of a rise and fall in aerosol emissions using a climate model ensemble forced by a sudden introduction of 1980s level sulfate aerosol emissions for 30 y, followed by a step function decrease to pre industrial sulfate aerosol emissions for the following 30 y. The initial increase in aerosol emissions leads to a fast warming, followed by a slow cooling, in the equatorial Pacific. The step function decline in aerosol emissions has the opposite effect—a fast cooling followed by a slow warming. In combination, these responses lead to a cooling effect in the equatorial Pacific that peaks after aerosols emissions are removed and persists for several decades (Fig. 1). A more realistic ensemble experiment, consisting of time-evolving aerosol emissions, illustrates a similar cooling with La Niña-like conditions emerging in the 1990s and lasting for several decades.

Fig. 1. The step function sulfate aerosol emission forcing considered by Hwang et al., with the dual timescale equatorial Pacific SST response below. Abrupt increase in aerosol emissions leads to a fast warming followed by a slow cooling. Abrupt decrease in aerosol emissions has the opposite effect. This dual timescale response allows for constructive combination and a surprisingly long-lasting cooling effect.

A new study in PNAS shows that an increase followed by a decrease in anthropogenic sulfate aerosol emissions can produce a multidecadal cooling pattern in the eastern tropical Pacific.

The fast response to anthropogenic sulfate aerosol emissions occurs over the first 3 y following the increase in aerosols. The wedge-shaped SST warming is driven by the formation of the cross-equatorial Hadley Cell. This fast response to aerosol changes hinges on low cloud-SST feedbacks and rainband location, which are commonly biased in global climate models. The slow response to sulfate aerosol emissions begins several years after aerosol emissions increase. The subsurface signature of the slow response continues for a decade after the removal of aerosols. In the cold tongue, this anomalously cold subsurface water is upwelled and kicks off a series of air–sea interactions which sustain a La Niña-like condition. Adequately simulating the slow response requires accurate low cloud-SST feedbacks and Pacific Ocean stratification. The model used by Hwang et al. has a realistic Pacific mean state and low cloud-SST feedbacks (10, 11) suggesting that it likely performs better than most models at producing this dual timescale SST response to changes in aerosols.

The response to the step function aerosol emission removal is the opposite of the response to aerosol emission increase—a fast cooling followed by a slow warming. The maximum cooling in the step function model ensemble occurs after the aerosol removal, when the slow response to the aerosol increase adds constructively to the fast cooling response from the aerosol removal (Fig. 1). Because the real world experienced a more linear aerosol emission increase until 1980, followed by an aerosol emission decrease, the expected response is a persistent subsurface cooling beginning in the 1960s and continuing through at least 2030. Reanalysis subsurface temperatures show a persistent cooling similar to that predicted by the model ensemble, supporting the existence of this mechanism in the real ocean.

A key component of Hwang et al.’s analysis is the use of single-model initial condition large ensembles to build a process-based understanding of anthropogenically forced climate responses. Any single model run contains random internal variability. Single-model initial condition large ensembles consist of a sufficient number of different ensemble members with different randomly phased internal variability, such that the mean of the ensemble members isolates the externally forced signal. These large ensembles are relatively new in climate science, due to the large amount of computing resources required. Recently, targeted spin-offs of single model initial condition large ensembles have been used to isolate specific processes, such as by comparing ensembles with different oceanic processes (12, 13) and analyzing large ensembles run under a single forcing scenario (such as time-varying greenhouse gases, anthropogenic aerosols, or biomass burning alone, with the other forcings held constant) (14, 15). Hwang et al. demonstrate how large ensemble experiments can push the boundaries of knowledge in climate science, by isolating and describing a previously unknown response of the tropical Pacific to time-evolving aerosol emissions.

Hwang et al.’s finding that time-evolving aerosol emissions are key to simulating a cooling in the eastern tropical Pacific presents a possible explanation for the disagreement between observed and long-term expected CO2 increase–driven radiative feedbacks. Since 1980, observed radiative feedbacks have been uncorrelated with the expected feedbacks from abrupt 4xCO2 forced model feedbacks, due to the unexpected pattern of tropical Pacific SST trends (16). Different spatial patterns of warming lead to different radiative feedbacks—in other words, the sensitivity of climate to a change in CO2 depends on the geographical pattern of induced warming. The tropical Pacific plays a dominant role in this effect. Warming in the western tropical Pacific is easily communicated upward and globally. When this is paired with less warming or cooling in the eastern tropical Pacific, extensive low clouds develop and serve as an efficient negative feedback to warming. Models generally suggest that increases in CO2 will give the opposite pattern, with a stronger warming in the eastern tropical Pacific compared to the west (2). This would reduce cloud cover and lead to less efficient cooling, or a less negative feedback. Hwang et al.’s work suggests that the relatively more stabilizing pattern effect from the recent historical period may be caused by time evolving aerosol emissions. This would imply that the recent time period is not very useful for constraining equilibrium climate sensitivity, because aerosols played a large role in the pattern effect, and not greenhouse gases alone. Taking this important role of aerosols into account may aid in recent efforts toward emergent constraint-based estimates of equilibrium climate sensitivity (17).

Crucially, Hwang et al. suggest that we may be nearing the peak of the aerosol-caused subsurface cooling, with their model suggesting that the peak cooling will occur between 2030 and 2050. If this hypothesis holds, we may soon begin to experience a shift toward a more El Niño-like warming pattern. This could have far-reaching impacts on climate, including the rate of global surface warming (16), droughts over the western United States (18), and changes to El Niño-Southern Oscillation (19). However, this timing likely depends on future aerosol emissions, which are assumed to continue to decline in the model used by Hwang et al. In reality, future aerosol emissions will be dictated by air-quality policies, and a sustained global decline is not guaranteed. Climate policy for the coming decades needs to consider the complex role that changes in sulfate aerosol emissions play. Continued research toward process-based understanding of climate responses to different time-evolving forcings, like Hwang et al.’s analyses, is an integral component of this effort.

Author contributions

K.M. wrote the paper.

Competing interests

The author declares no competing interest.

See companion article, “Contribution of anthropogenic aerosols to persistent La Niña-like conditions in the early 21st century,” 10.1073/pnas.2315124121.
==== Refs
1 J. E. Kay , The community earth system model (CESM) large ensemble project: A community resource for studying climate change in the presence of internal climate variability. Bull. Am. Meteorol. Soc. 96 , 1333–1349 (2015).
2 R. C. J. Wills, Y. Dong, C. Proistosecu, K. C. Armour, D. S. Battisti, Systematic climate model biases in the large-scale patterns of recent sea-surface temperature and sea-level pressure change. Geophys. Res. Lett. 49 , e2022GL100011 (2022).
3 R. Seager, N. L. Henderson, M. A. Cane, Persistent discrepancies between observed and modeled trends in the tropical Pacific Ocean. J. Clim. 35 , 4571–4584 (2022).
4 Y.-T. Hwang, S.-P. Xie, P.-J. Chen, H.-Y. Tseng, C. Deser, Contribution of anthropogenic aerosols to persistent La Niña-like conditions in the early 21st century. Proc. Natl. Acad. Sci. U.S.A. 121 , e2315124121 (2024).38252827
5 J.-F. Lamarque , Historical (1850–2000) gridded anthropogenic and biomass burning emissions of reactive gases and aerosols: Methodology and application. Atmos. Chem. Phys. 10 , 7017–7039 (2010).
6 R. M. Hoesly , Historical (1750–2014) anthropogenic emissions of reactive gases and aerosols from the Community Emissions Data System (CEDS). Geosci. Model Dev. 11 , 369–408 (2018).
7 J. Hansen, M. Sato, R. Ruedy, Radiative forcing and climate response. J. Geophys. Res.: Atmos. 102 , 6831–6864 (1997).
8 N. Bellouin , Bounding global aerosol radiative forcing of climate change. Rev. Geophys. 58 , e2019RG000660 (2020).
9 U. K. Heede, A. V. Fedorov, Eastern equatorial Pacific warming delayed by aerosols and thermostat response to CO2 increase. Nat. Clim. Change 11 , 696–703 (2021).
10 H. Kim, S. M. Kang, J. E. Kay, S.-P. Xie, Subtropical clouds key to Southern Ocean teleconnections to the tropical Pacific. Proc. Natl. Acad. Sci. U.S.A. 119 , e2200514119 (2022).35969773
11 Y. Y. Planton , Evaluating climate models with the CLIVAR 2020 ENSO metrics package. Bull. Am. Meteorol. Soc. 102 , E193–E217 (2021).
12 L. N. Murphy, J. M. Klavans, A. C. Clement, M. A. Cane, Investigating the roles of external forcing and ocean circulation on the Atlantic multidecadal SST variability in a large ensemble climate model hierarchy. J. Clim. 34 , 4835–4849 (2021).
13 K. McMonigal, S. Larson, S. Hu, R. Kramer, Historical changes in wind-driven ocean circulation can accelerate global warming. Geophys. Res. Lett. 50 , e2023GL102846 (2023).
14 I. R. Simpson , The CESM2 single-forcing large ensemble and comparison to CESM1: Implications for experimental design. J. Clim. 36 , 5687–5711 (2023).
15 C. Deser , Isolating the evolving contributions of anthropogenic aerosols and greenhouse gases: A new CESM1 large ensemble community resource. J. Clim. 33 , 7835–7858 (2020).
16 T. Andrews , On the effect of historical SST patterns on radiative feedback. J. Geophys. Res. Atmos. 127 , e2022JD036675 (2022).
17 S. C. Sherwood , An assessment of earth’s climate sensitivity using multiple lines of evidence. Rev. Geophys. 58 , 1–92 (2020).
18 R. Seager , Causes of the 2011–14 California drought. J. Clim. 28 , 6997–7024 (2015).
19 W. Cai , Changing El Niño-Southern oscillation in a warming climate. Nat. Rev. Earth Environ. 2 , 628–644 (2021).
