the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 12 Nov 2025
The global climate response to High-Latitude Low-Altitude Stratospheric Aerosol Injection (HiLLA-SAI)
Abstract. High-latitude low-altitude (HiLLA) Stratospheric Aerosol Injections (SAI) would face fewer logistical barriers than high-altitude low-latitude SAI, because it could use repurposed existing large aircraft for deployment. However, relative to high-altitude SAI, it is expected to have reduced global cooling efficiency, and the more polar forcing profile and reduced tropical stratospheric heating would result in many differences in the surface climate response. Here, we present the first multi-model simulations of HiLLA-SAI, in UKESM1, CESM2-WACCM and E3SMv3. Using these simulations, we assess the global climate response to HiLLA-SAI, and the sensitivity to the latitude, altitude (13 km versus 15 km), seasonality and longitude of injections. For seasonal injections at 60° N/S and 13 km, all models show similar global cooling efficiency, of around 0.6 °C per 12 Mt SO2 per year, 40–53 % of the equivalent cooling efficiency for 21 km injection in the tropics. Raising the injection height to 15 km increases this global cooling efficiency by around half, to 63–70 % of the high altitude tropical case. The effects of HiLLA-SAI are more polar focused than other SAI strategies, particularly for the 13 km injection case, and large changes in sea-ice in both hemispheres, high-latitude precipitation and the polar seasonal cycle are shown. Nevertheless, our results highlight that HiLLA-SAI would still be a global intervention. For 13 km inject, tropical cooling per unit global cooling is 61–75 % of the rate under greenhouse-gas forced warming, and is larger in the 15 km case. Precipitation changes and sulfur deposition are also found at all latitudes. Overall, our results highlight the importance of further study into HiLLA-SAI strategies, which these simulations suggest could be a viable early-stage SAI deployment strategy, with global, not just polar, impacts.
- Preprint
(22085 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-5356', Anonymous Referee #1, 21 Nov 2025
-
AC1: 'Reply on RC1', Alistair Duffey, 06 Mar 2026
We thank the reviewer for their helpful comments, which will be very useful in revising the manuscript. Please see below for our planned changes in response to each comment:
Line 12 – “For 13 km inject,” -> “For the 13 km injection,”Thanks, we have fixed this typo.
Line 12-13 – This sentence reads awkward to me. I’m not sure I completely follow what is meant by “tropical cooling per unit global cooling” under a GHG warming scenario. Why is there cooling in a warming scenario – perhaps I just don’t understand this metric.
We agree this was confusingly phrased. We will revise sentence this to read as follows: "tropical temperature change per unit global temperature change is 61-75% of the equivalent ratio under greenhouse-gas forced warming"
Line 22 - The phrase “well above … quantities of SO2” could use a citation (Smith et al., 2024?)
Agreed, we will cite Smith et al. (2024) here, as suggested.
Line 45 – Exchange the “;” for a “.”
Agreed, we will make this edit.
Line 48 – Change “68N, 10-15 km” to “68N at 10-15 km”?
Agreed, we will make this edit.
Line 100 – citation for CAM
Thanks, on second thoughts we will remove this reference to CAM as potentially confusing given our simulations use only WACCM. So the sentence would now read: The atmospheric component used for these simulations is the Whole Atmosphere Community Climate Model (WACCM6; Gettelman et al., 2019), which uses a finite-volume dynamical core and uses 70 vertical levels, up to approximately 140 km.
In Section 2.2, note somewhere that there is only 1 ensemble member for each HiLLA experiment and that they branch from the SSP2-4.5 r1. … I see “r1” written in some figures and figure captions, but I don’t believe this term is explained in the text.
Thanks, we will add this information by including the following sentence at the end of section 2.2:
"A single ensemble member is run for each simulation, and all simulations branch from the same member of SSP2-4.5. In UKESM1 this SSP2-4.5 simulation is ‘r1i1p1f2’, in CESM2-WACCM it is ‘r1i1p1f1’, and in E3SMv3 a novel control simulation following the SSP2-4.5 scenario was simulated. We refer to this background run as ‘r1’ in each case. For UKESM1 and CESM2-WACCM, we also show the ensemble range and mean in figures below, but in E3SMv3 only a single control member is available."
Figure 3 caption – last sentence: How is the ‘Antarctic’ defined?
Thanks for highlighting this omision - we will add a definition for the Antarctic area as follows: "The `Tropics', `Arctic' and `Antarctic' are defined as (23°S-23°N), >66°N, and >66°S, respectively."
Line 192 – remove “,” after the word cooling
Thanks, we will remove the comma.
Line 223-224 – Could you explain the purpose of this decomposition, I’m not sure I follow the goal of this analysis
Thanks, yes, we will add additional explanation of the purpose here as follows: "We now decompose the global cooling per unit injection into two components: the global mean AOD per unit injection, and the global mean cooling per unit global mean AOD. This separates the effects of variation in aerosol lifetime and size distribution (which define the AOD per unit injection) from the effects of variation in the combination of SW forcing per unit AOD and temperature sensitivity to that global mean forcing, which together define cooling per unit global mean AOD (Kleinschmitt et al., 2018; Zhang et al., 2024; Duffey et al., 2025a)."
Line 309 – Specify which relationship is “weaker”
Agreed - we will revise the sentence to specify as follows: "and the relationship between local annual mean temperature and sea-ice area is weaker."
Figure 14 Caption – last sentence, CESM_WACCM should be CESM-WACCM
Thanks, fixed. We also note (as pointed out by the 2nd reviewer) that there were a few inconsistencies throughout the manuscript in model naming (e.g. ‘CESM’, ‘CESM2’, and ‘CESM2-WACCM’ where used in different places) - the revised manuscript will use ‘UKESM1’, ‘CESM2-WACCM’ and ‘E3SMv3’ throughout.
…
References
Smith, W., Bartels, M. F., Boers, J. G., & Rice, C. V. (2024). On thin ice: Solar geoengineering to manage tipping element risks in the cryosphere by 2040. Earth's Future, 12, e2024EF004797. https://doi. org/10.1029/2024EF004797
Citation: https://doi.org/10.5194/egusphere-2025-5356-AC1
-
AC1: 'Reply on RC1', Alistair Duffey, 06 Mar 2026
-
RC2: 'Comment on egusphere-2025-5356', Anonymous Referee #2, 06 Feb 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-5356/egusphere-2025-5356-RC2-supplement.pdf
- AC2: 'Reply on RC2', Alistair Duffey, 06 Mar 2026
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 837 | 208 | 38 | 1,083 | 41 | 36 |
- HTML: 837
- PDF: 208
- XML: 38
- Total: 1,083
- BibTeX: 41
- EndNote: 36
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Summary:
The authors examine a specific SAI strategy that departs from the standard tropical, high-altitude injection approach. Instead, they focus on high-latitude, low-altitude injections (HiLLA-SAI), where SO₂ is released near 60° N/S at 13–15 km. The motivation is that these lower-altitude, high-latitude injections could be carried out with existing large aircraft, making them a potentially more practical option for an early-stage deployment. Using simulations from three Earth system models (UKESM1, CESM2-WACCM, and E3SMv3), the authors evaluate how variations in injection altitude, latitude, season, and longitude influence the climate response. They find that HiLLA-SAI is less efficient per unit SO₂ than traditional tropical high-altitude SAI, but it still produces notable global cooling and strong high-latitude impacts, including changes in sea ice, polar precipitation, and the seasonal cycle. Although the forcing is concentrated at high latitudes, the authors show that HiLLA-SAI nonetheless drives global-scale responses—affecting tropical temperatures, precipitation patterns, and sulfur deposition—and therefore should not be viewed as a purely regional intervention. They conclude by emphasizing the need for continued study of such strategies, particularly given their potential relevance for near-term or limited-capacity deployment scenarios.
Minor Comments:
Line 12 – “For 13 km inject,” -> “For the 13 km injection,”
Line 12-13 – This sentence reads awkward to me. I’m not sure I completely follow what is meant by “tropical cooling per unit global cooling” under a GHG warming scenario. Why is there cooling in a warming scenario – perhaps I just don’t understand this metric.
Line 22 - The phrase “well above … quantities of SO2” could use a citation (Smith et al., 2024?)
Line 45 – Exchange the “;” for a “.”
Line 48 – Change “68N, 10-15 km” to “68N at 10-15 km”?
Line 100 – citation for CAM
In Section 2.2, note somewhere that there is only 1 ensemble member for each HiLLA experiment and that they branch from the SSP2-4.5 r1. … I see “r1” written in some figures and figure captions, but I don’t believe this term is explained in the text.
Figure 3 caption – last sentence: How is the ‘Antarctic’ defined?
Line 192 – remove “,” after the word cooling
Line 223-224 – Could you explain the purpose of this decomposition, I’m not sure I follow the goal of this analysis
Line 309 – Specify which relationship is “weaker”
Figure 14 Caption – last sentence, CESM_WACCM should be CESM-WACCM