CoChat

Air quality impacts of stratospheric aerosol injections are likely small and mainly driven by changes in climate, not aerosol settling

Open access
Wang, Cindy, Visioni, Daniele, Chua, Glen, Bednarz, Ewa M. · 2026 · Atmospheric Chemistry and Physics

Published version of a preprint

The DOI matches its Crossref record; 89 of the 89 references it cites were found in a registry or an index; 34 of the 79 sources its claims cite were checked for support.

Abstract

Stratospheric aerosol injection (SAI) is a proposed climate intervention method to offset future global warming through increased solar reflection in the stratosphere, but its broader environmental and public health implications are yet to be thoroughly explored. We use three large ensembles of fully coupled CESM2-WACCM6 simulations to assess changes in mortality attributable to fine particulate matter (PM2.5) and surface ozone exposure (O3). Maintaining temperatures at 1.5 °C above preindustrial levels through SAI is projected to yield a modest 0.4 % (ensemble range: −1.9 % to +1.5 %) reduction in pollution-related mortality relative to middle-of-the-road climate change scenario, reflecting a 1.3 % (−2.3 % to −0.6 %) reduction in ozone-related deaths and a 0.9 % (−0.4 % to +2.1 %) increase in PM2.5-related deaths. The spread among ensemble members underscores the influence of internal variability and highlights the importance of ensemble-based analyses when assessing the potential health impacts of climate intervention strategies. We find that global PM2.5 mortality changes exhibit little sensitivity to injected sulfate amounts, with the most variability driven by precipitation-mediated changes in non-sulfate PM2.5 species (e.g., dust and secondary organic aerosols), whereas ozone-related mortality is primarily driven by surface cooling and hemispheric asymmetries in stratospheric-tropospheric exchange and ozone transport. However, our results heavily reflect the specific forcing patterns of the SAI scenarios used; our estimates are also limited by model shortcomings, including omitting the effects of aerosols in the photolysis scheme – which might limit UV-driven changes and impact surface ozone rates – or not including nitrate aerosols. Within our framework, we find that SAI impacts on pollution-related mortality are modest but regionally heterogeneous, and that the magnitude of the SAI-driven changes is smaller than the improvements expected from near-term air quality policies planned or implemented within the same time frame.

Overview

CoChat Summary

TL;DR

Using three 10-member CESM2-WACCM6 ensembles, this study estimates that stratospheric aerosol injection (SAI) to hold warming at 1.5 °C produces only modest changes in air-pollution mortality relative to SSP2-4.5. The PM2.5 mortality response is driven mainly by climate-mediated changes in non-sulfate species, while ozone mortality reflects hemispheric asymmetries in stratospheric-tropospheric exchange and tropospheric chemistry.

Key findings

  • SAI yields a 0.4 % reduction in total pollution-related mortality relative to SSP2-4.5, reflecting a 1.3 % reduction in ozone-related deaths and a 0.9 % increase in PM2.5-related deaths.
  • Absolute estimates: SAI leads to a net increase of 14 000 PM2.5-related premature deaths and a net reduction of approximately 14 000 ozone-related premature deaths compared with SSP2-4.5 over 2060–2069.
  • PM2.5 mortality changes are not driven by sulfate; dust and secondary organic aerosols dominate, with precipitation and circulation affecting wet removal. Sulfate’s contribution is small relative to other components.
  • Ozone mortality reductions occur mainly in the Southern Hemisphere from reduced stratosphere-troposphere exchange, while Northern Hemisphere surface ozone increases due to drier, colder tropospheric conditions suppressing photochemical ozone loss.
  • SAI’s net effect is small compared with policy-driven improvements: SSP2-4.5 reduces air-pollution mortality by 18 % relative to present day, and ARISE-SAI-1.5 by 19 %, with the SAI impact largely within internal variability.

Methods

  • Fully coupled CESM2-WACCM6 simulations with interactive aerosols and chemistry: three 10-member ensembles—SSP2-4.5 baseline, ARISE-SAI-1.5, and ARISE-SAI-1.0. SAI injects SO2 at 15° N, 15° S, 30° N, and 30° S at approximately 21.5 km from 2035 to 2070.
  • Mortality estimated with a health impact function using a fixed 2020 population distribution; PM2.5 exposure-response from the Global Exposure Mortality Model, ozone from ozone season maximum daily 8 h average with GBD risk functions, and baseline mortality rates from the International Futures health model.
  • Analyses compare 2060–2069 with 2030–2039 and isolate SAI by comparing ARISE-SAI-1.5 with SSP2-4.5 over the same future period; ensemble spread is used to characterize internal variability.

Limitations

  • Stated by the authors: photolysis rates use lookup tables that exclude aerosol effects; the model omits ammonium and nitrate aerosols; dust concentrations may be overestimated; mortality uses a fixed 2020 population; results are based on a single model and two closely related SAI scenarios with SO2 injected primarily in the Southern Hemisphere.
  • Also stated: other health-relevant outcomes such as surface UV and regional temperature changes were not evaluated, and processes such as planetary boundary layer height and lightning activity were not included.
  • Evident from the text: large ensemble spread means regional mortality changes can differ in sign, and the results are scenario- and model-dependent.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.5194/acp-26-1339-2026

Cites

89 of 89 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Spatial patterns of the changes in surface climate variables (surface temperature and precipitation), air quality (PM 2 : 5 and OSMDA8) under the SAI scenario (ARISE-SAI-1.5) and the baseline scenario (SSP2-4.5) for the period 2060-2069 compared to 2030-2039. Each row represents changes in surface temperature ( T s , K) (a-c) , precipitation (mm d 1 ) (d-f) , PM2 : 5 concentration (µg m 3 ) (g-i) , and OSMDA8 (ppb) (j-1) . The stippling indicates areas where differences between ARISE-SAI-1.5 and SSP2-4.5 are not statistically significant ( p> 0.05) based on a t -test performed across all 10 ensemble members. Columns indicate the difference between the SAI case and the reference period with same global temperatures (left), the difference between a warmer future and the reference period (center), and the difference between the SAI case and a warmer future following the same underlying emission scenario (right).
Figure 1 Spatial patterns of the changes in surface climate variables (surface temperature and precipitation), air quality (PM 2 : 5 and OSMDA8) under the SAI scenario (ARISE-SAI-1.5) and the baseline scenario (SSP2-4.5) for the period 2060-2069 compared to 2030-2039. Each row represents changes in surface temperature ( T s , K) (a-c) , precipitation (mm d 1 ) (d-f) , PM2 : 5 concentration (µg m 3 ) (g-i) , and OSMDA8 (ppb) (j-1) . The stippling indicates areas where differences between ARISE-SAI-1.5 and SSP2-4.5 are not statistically significant ( p> 0.05) based on a t -test performed across all 10 ensemble members. Columns indicate the difference between the SAI case and the reference period with same global temperatures (left), the difference between a warmer future and the reference period (center), and the difference between the SAI case and a warmer future following the same underlying emission scenario (right).From the PDF
Figure 2 Map of the most prevalent PM 2 : 5 species (dust, POM, salt, SO 4 , SOA and BC) across grid cells, derived from ensemble model averages under the ARISE-SAI-1.5 scenario. Colors represent the dominant species at each location, determined by taking the fraction of the species to the total PM 2 : 5 concentration. Black carbon is not presented here because it does not dominate in any grid cell. White stippling is over areas where fewer than 90 % of ensemble members agree on the dominant species at a grid point.
Figure 2 Map of the most prevalent PM 2 : 5 species (dust, POM, salt, SO 4 , SOA and BC) across grid cells, derived from ensemble model averages under the ARISE-SAI-1.5 scenario. Colors represent the dominant species at each location, determined by taking the fraction of the species to the total PM 2 : 5 concentration. Black carbon is not presented here because it does not dominate in any grid cell. White stippling is over areas where fewer than 90 % of ensemble members agree on the dominant species at a grid point.From the PDF
Image 3
From the PDF
Image 4
From the PDF
Image 5
From the PDF
Figure 6 Country-level ozone-related mortality changes (per 100 000 people) for ARISE-SAI-1.5 and SSP2-4.5 simulations. Each row shows the ensemble-mean mortality change (left column) and the corresponding inter-ensemble standard deviation (right column) across ten CESM2(WACCM6) ensemble members. Row (a) -(b) : ARISE-SAI-1.5 for 2060-2069 relative to SSP2-4.5 for 2030-2039; (c) -(d) : SSP24.5 for 2060-2069 relative to its own 2030-2039 baseline; (e) -(f) : ARISE-SAI-1.5 relative to SSP2-4.5 for the same 2060-2069 period, isolating the direct SAI effect. Mortality values are normalized by total population in each country and expressed in deaths per 100 000 people. Stippling indicates countries where estimates of the ozone-related mortality are not statistically significant across ensemble members at the 95 % confidence level.
Figure 6 Country-level ozone-related mortality changes (per 100 000 people) for ARISE-SAI-1.5 and SSP2-4.5 simulations. Each row shows the ensemble-mean mortality change (left column) and the corresponding inter-ensemble standard deviation (right column) across ten CESM2(WACCM6) ensemble members. Row (a) -(b) : ARISE-SAI-1.5 for 2060-2069 relative to SSP2-4.5 for 2030-2039; (c) -(d) : SSP24.5 for 2060-2069 relative to its own 2030-2039 baseline; (e) -(f) : ARISE-SAI-1.5 relative to SSP2-4.5 for the same 2060-2069 period, isolating the direct SAI effect. Mortality values are normalized by total population in each country and expressed in deaths per 100 000 people. Stippling indicates countries where estimates of the ozone-related mortality are not statistically significant across ensemble members at the 95 % confidence level.From the PDF
Image 7
From the PDF
Figure 8 Global mortality differences for ARISE-SAI-1.5 minus SSP2-4.5 (shades of green) and ARISE-SAI-1.0 minus SSP2-4.5 (shades of blue) as a function of annual SO 2 injection rate (Tg-SO 2 yr 1 ), for PM 2 : 5 -related deaths (green) and ozone-related deaths (green). Points represent ensemble means across years (2035-2069), with error bars indicating ensemble variability ( 1 standard deviation).
Figure 8 Global mortality differences for ARISE-SAI-1.5 minus SSP2-4.5 (shades of green) and ARISE-SAI-1.0 minus SSP2-4.5 (shades of blue) as a function of annual SO 2 injection rate (Tg-SO 2 yr 1 ), for PM 2 : 5 -related deaths (green) and ozone-related deaths (green). Points represent ensemble means across years (2035-2069), with error bars indicating ensemble variability ( 1 standard deviation).From the PDF
Figure 9 (a) Global % change in mortality burden comparing ARISE-SAI-1.5 (2060-2069) and SSP2-4.5 (2030-2039), as well as SSP2-4.5 (2060-2069) and SSP2-4.5 (2030-2039). (b) Absolute global changes in mortality burden (in number of deaths). (c) %change in mortality burden by Global Burden of Disease (GBD) super-region between ARISE-SAI-1.5 (2060-2069) and SSP2-4.5 (2060-2069). % changes are calculated relative to baseline mortality rates. Positive values indicate an increase in mortality relative to the baseline, while negative values indicate reductions. Bars represent stacked contributions from ozone-related deaths (blue) and PM 2 : 5 -related deaths (orange), with horizontal error bars indicating the ensemble spread (standard deviation) for each component and for the net total (black diamonds with error bars).
Figure 9 (a) Global % change in mortality burden comparing ARISE-SAI-1.5 (2060-2069) and SSP2-4.5 (2030-2039), as well as SSP2-4.5 (2060-2069) and SSP2-4.5 (2030-2039). (b) Absolute global changes in mortality burden (in number of deaths). (c) %change in mortality burden by Global Burden of Disease (GBD) super-region between ARISE-SAI-1.5 (2060-2069) and SSP2-4.5 (2060-2069). % changes are calculated relative to baseline mortality rates. Positive values indicate an increase in mortality relative to the baseline, while negative values indicate reductions. Bars represent stacked contributions from ozone-related deaths (blue) and PM 2 : 5 -related deaths (orange), with horizontal error bars indicating the ensemble spread (standard deviation) for each component and for the net total (black diamonds with error bars).From the PDF

Verification

Claims the paper makes

22 Valid · 6 Not valid · 16 Not checked · 86 Own results and interpretations

55 more claims · 2 with their cited sources checked

Create a free account to see all 130 claims and to verify the 16 references still open.

See all claims — free account
  • Probably not a substantial claim. 33
  • It may state two separate claims as one. 50
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

4 of 4

Cited works that were retracted

Works this paper cites that were later retracted, from the retraction notices publishers deposit. Major when the notice came out before this paper was published.

1 of 1

Citations

Cites

1 retracted

Show all 89 references

Cited by

See all 3

Metadata

Authors
Wang, C., Visioni, D., Chua, G., & Bednarz, E. M.
Affiliations
  • Cornell University (US)
  • Goddard Institute for Space Studies (US)
  • Cooperative Institute for Research in Environmental Sciences (US)
Corresponding author
Wang, Cindy
Venue
Atmospheric Chemistry and Physics
Published
2026-01-27
Volume
26
Issue
2
Pages
1339-1357
ISSN
  • 1680-7316 (ISSN-L)
  • 1680-7324
OpenAlex
W7125783580
Cited by
3
References
78
Citations by year
2026: 3
FWCI
9.04
Type
journal article
Language
en
Open access
Locations
  • Atmospheric chemistry and physics (publishedVersion, cc-by)
  • NOAA Institutional Repository (submittedVersion, cc-by)
Funders
Quadrature Climate Foundation (N/A)
Topics
  • Physical Sciences › Environmental Science › Global and Planetary Change › Climate Change and Geoengineering
  • Physical Sciences › Environmental Science › Global and Planetary Change › Atmospheric aerosols and clouds
  • Physical Sciences › Earth and Planetary Sciences › Atmospheric Science › Atmospheric chemistry and aerosols
Keywords
stratospheric aerosol injection, solar geoengineering, climate intervention, air quality, mortality, PM2.5, surface ozone, ozone-related mortality, sulfate aerosols, stratosphere-troposphere exchange, hemispheric asymmetry (solar and geophysical), secondary organic aerosol, dust, public health impact, CESM2-WACCM6

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Air quality impacts of stratospheric aerosol injections are likely small and mainly driven by changes in climate, not aerosol settling. https://app.cochat.ai/papers/10.5194/acp-26-1339-2026
The paper
Wang, C., Visioni, D., Chua, G., & Bednarz, E. M. (2026). Air quality impacts of stratospheric aerosol injections are likely small and mainly driven by changes in climate, not aerosol settling. Atmospheric Chemistry and Physics. https://doi.org/10.5194/acp-26-1339-2026

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.