the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 27 May 2026
Estimating Twomey forcing sensitivity to aerosol plume spreading rates
Abstract. The amount of sunlight that reaches Earth's surface can be reduced by increasing cloud droplet number and decreasing droplet size (i.e., Twomey forcing), a central idea underpinning the Marine Cloud Brightening strategy. Cloud albedo depends nonlinearly on cloud droplet concentration, meaning the spatial extent of aerosol plumes could be an important constraint on the brightening potential. Horizontal spreading of aerosols can be simulated by Langevin particle models, which depend on turbulence characteristics and precipitation. The turbulence information that drives the particle model is derived from a library of realistic, 2-day LES cases, spanning northeast Pacific stratocumulus conditions. The 2-D reflectance fields from the LES are superimposed onto the 2-D perturbed aerosol concentrations to calculate the 2-D Twomey forcing response.
The Day 1 and Day 2 LES regimes have distinct meteorological, aerosol, and turbulence characteristics associated with equatorward movement of the LES domain. Our results indicate that the Day 2 regime has substantially faster plume spreading than Day 1, with ensemble median differences exceeding 2 km hr-1. Despite these differences, Twomey forcing is insensitive to the natural variability in spreading rate. This study suggests that Twomey forcing is resilient to variations in meteorology, aerosols, and turbulence, with its efficacy governed primarily by aerosol lifetimes and assumptions surrounding cloud adjustments. Although the natural variation in spreading rate does not materially affect Twomey forcing, idealized plume spreading simulations suggest that current GCM assumptions of an infinitely fast spreading rate could lead to a 10-200% overestimation of cooling.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(11672 KB)
-
Supplement
(1788 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(11672 KB) - Metadata XML
-
Supplement
(1788 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2026-2730', Anonymous Referee #1, 21 Jul 2026
-
AC1: 'Reply on RC1', Lucas McMichael, 04 Sep 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-2730/egusphere-2026-2730-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Lucas McMichael, 04 Sep 2026
-
RC2: 'Comment on egusphere-2026-2730', Anonymous Referee #2, 02 Aug 2026
Mcmichael et al 2026 ACP:This article investigates the brightening potential of an MCB like aerosol perturbation to the NEP marine boundary layer. The authors have cleverly combined LES simulations with an offline particle transport model to estimate the Twomey component of the cloud radiative effect. This is an extension of the Idealised Wood 2021 ACP paper and will be of interest to the MCB community. While the authors have explored several subtopics in this idealised study, it would have been better if they had tried to match some of their results with the existing LES studies. The data are all publicly available. This would add more credibility to their model before embarking on an exercise with GCM(s), which I belive is the next logical step. Additionally, it would be useful if the authors could trim certain portions of the results section to enhance the readability of the manuscript. As it stands, it is slightly on the longish end.My recommendation: publish after the comments below are addressed.Comments:1. Sec 2.1 and 2.2, can you attribute any physical significance to PC1 and PC2?2. Line 101: quantify "reasonably capture". What does it mean?3. Lines 146 - In LES you are not solving N-S equations. You are solving a filtered form of the N-S equations.4. Why is the LPM domain size substantially different from that of the LES?5. Lines 201 to 204: I am not able to understand what the authors are trying to say. Can you rephrase? And equilibrium in what sense?6. Line 221: Isn't Finc a constant in space? Then why the overbar?7. Lines 229-30: Is it SW radiative forcing or radiative effect? From what I understand it should be a radiative effect. Please clarify and check throughout the manuscript. Forcing is used repeatedly.8. Sec 3.1, how are the LWP and cloud fractions defined? Are there any thresholds on LWC or cloud optical depth?9. Line 317: How can they all influence in nonlinear ways? Can you explain this a bit? It was stated at multiple locations throughout the manuscript.10. Fig 7: is there a way to bring in some uncertainty estimates? Without that I am not sure how to interpret these numbers. Are all of these upper bounds?11. Lines 360-370: I am not able to follow how the susceptibility analysis was carried out. Can you please explain it better? Consequently, I am not able to appreciate the importance of Fig. 8.12. How important is the subgrid scale representation of plume tracks in GCMs considering the high uncertainty associated with ACI representation? Would including these subgrid scale effects reduce the uncertainty or increase the uncertainty from an MCB perspective?13. How would incorporating activation fraction change your results? Have a reservoir of aerosol and activate them when the need arises. This probably is more closer to reality than the current treatment.14. Line 453: shouldn't it be divided by TL (TKE/TL)?15. The results here seem to suggest that the timing of the aerosol injection is important. Can you explain this a bit? Is it coming from the role played by precipitation? Prabhakaran et al 2023 and 2024 seems to indicate that timing of the aerosol injection doesn't matter for non-precip. clouds. However, Zhang et al 2023 looked at an ensemble of LES studies and concluded that timing matters even in non-precip clouds. Thoughts?Citation: https://doi.org/
10.5194/egusphere-2026-2730-RC2 -
AC2: 'Reply on RC2', Lucas McMichael, 04 Sep 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-2730/egusphere-2026-2730-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Lucas McMichael, 04 Sep 2026
Peer review completion
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2026-2730', Anonymous Referee #1, 21 Jul 2026
Lucas et al. present an interesting study in which they use Langevin particle modelling applied to realistic LES outputs to investigate the importance of the emission spreading rate of sprayers (ships) for Twomey forcing. They further examine the influence of variability in aerosol properties, including the number of sprayers and their motion type, aerosol decay timescales and age, concentration bins, and background aerosol concentrations, as well as meteorological conditions. Their results show that the velocity and type of sprayer motion have little influence on Twomey forcing. Assuming that cloud adjustments are relatively small, they further demonstrate that the effect of the spreading rate on Twomey forcing is limited, with the forcing being strongly controlled by aerosol lifetime. Their findings suggest that the common assumption in GCMs of an infinitely fast spreading rate could result in a 10–200% overestimation of Twomey forcing. Overall, I found the manuscript to be well written, and the methodology to be sound. I have only a few minor comments that can be addressed easily by the authors.
Minor comments:
1. Line 43: Goren et al. (2025) also attributed the co-variability to two factors: (1) large-scale meteorological conditions, which influence cloud properties such as cloud liquid water path (LWP) and cloud droplet number concentration (Nd), causing these variables to covary simultaneously and inversely; and (2) microphysical processes associated with cloud development, during which an increase in LWP is generally accompanied by a decrease in Nd.
Goren et al. (2025): Co-variability drives the inverted-V sensitivity between liquid water path and droplet concentrations, Atmos. Chem. Phys., 25, 3413–3423. https://doi.org/10.5194/acp-25-3413-2025
2. Line 45: By constraining cloud morphology using MODIS observations, Goren et al. (2026) reported contrasting results, finding that LWP adjustments nearly offset the Twomey forcing.
Goren et al (2026): Beyond discrete stratocumulus regimes: a ternary continuum of morphology reveals within-regime variability in cloud susceptibilities, Atmos. Chem. Phys., 26, 7193–7206. https://doi.org/10.5194/acp-26-7193-2026
3. Line 75–76: Are there any references demonstrating that these 17 LES simulations are in close agreement with the observations. Or the term “observations” is used loosely here to represent general shallow marine clouds. Please clarify.
4. Line 76: The abbreviation LPM appears for the first time here without being defined.
5. Line 118: Could the authors also describe how Nd is derived from the CERES measurements? In addition, how is Na computed from the MERRA-2 data? A brief description or an appropriate reference would be sufficient.
6. Line 263: This assumption is inconsistent with Equation (2), where only 80% of aerosol particles, rather than 100%, are assumed to activate as cloud droplets.
7. Line 272: Please define NAc
8. Figure 3: Please define what the top panel in each panel group represents in the figure caption.
9. Line 391 and Figure 10 (also in other instances): Is the twomey forcing calculated for these simulations “global”?
10. Section 4: Could the authors also comment on the role of cloud morphology and cell size on the spreading rate and Twomey forcing relationship?
Citation: https://doi.org/10.5194/egusphere-2026-2730-RC1 -
AC1: 'Reply on RC1', Lucas McMichael, 04 Sep 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-2730/egusphere-2026-2730-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Lucas McMichael, 04 Sep 2026
-
RC2: 'Comment on egusphere-2026-2730', Anonymous Referee #2, 02 Aug 2026
Mcmichael et al 2026 ACP:This article investigates the brightening potential of an MCB like aerosol perturbation to the NEP marine boundary layer. The authors have cleverly combined LES simulations with an offline particle transport model to estimate the Twomey component of the cloud radiative effect. This is an extension of the Idealised Wood 2021 ACP paper and will be of interest to the MCB community. While the authors have explored several subtopics in this idealised study, it would have been better if they had tried to match some of their results with the existing LES studies. The data are all publicly available. This would add more credibility to their model before embarking on an exercise with GCM(s), which I belive is the next logical step. Additionally, it would be useful if the authors could trim certain portions of the results section to enhance the readability of the manuscript. As it stands, it is slightly on the longish end.My recommendation: publish after the comments below are addressed.Comments:1. Sec 2.1 and 2.2, can you attribute any physical significance to PC1 and PC2?2. Line 101: quantify "reasonably capture". What does it mean?3. Lines 146 - In LES you are not solving N-S equations. You are solving a filtered form of the N-S equations.4. Why is the LPM domain size substantially different from that of the LES?5. Lines 201 to 204: I am not able to understand what the authors are trying to say. Can you rephrase? And equilibrium in what sense?6. Line 221: Isn't Finc a constant in space? Then why the overbar?7. Lines 229-30: Is it SW radiative forcing or radiative effect? From what I understand it should be a radiative effect. Please clarify and check throughout the manuscript. Forcing is used repeatedly.8. Sec 3.1, how are the LWP and cloud fractions defined? Are there any thresholds on LWC or cloud optical depth?9. Line 317: How can they all influence in nonlinear ways? Can you explain this a bit? It was stated at multiple locations throughout the manuscript.10. Fig 7: is there a way to bring in some uncertainty estimates? Without that I am not sure how to interpret these numbers. Are all of these upper bounds?11. Lines 360-370: I am not able to follow how the susceptibility analysis was carried out. Can you please explain it better? Consequently, I am not able to appreciate the importance of Fig. 8.12. How important is the subgrid scale representation of plume tracks in GCMs considering the high uncertainty associated with ACI representation? Would including these subgrid scale effects reduce the uncertainty or increase the uncertainty from an MCB perspective?13. How would incorporating activation fraction change your results? Have a reservoir of aerosol and activate them when the need arises. This probably is more closer to reality than the current treatment.14. Line 453: shouldn't it be divided by TL (TKE/TL)?15. The results here seem to suggest that the timing of the aerosol injection is important. Can you explain this a bit? Is it coming from the role played by precipitation? Prabhakaran et al 2023 and 2024 seems to indicate that timing of the aerosol injection doesn't matter for non-precip. clouds. However, Zhang et al 2023 looked at an ensemble of LES studies and concluded that timing matters even in non-precip clouds. Thoughts?Citation: https://doi.org/
10.5194/egusphere-2026-2730-RC2 -
AC2: 'Reply on RC2', Lucas McMichael, 04 Sep 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-2730/egusphere-2026-2730-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Lucas McMichael, 04 Sep 2026
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 218 | 134 | 54 | 406 | 53 | 39 | 36 |
- HTML: 218
- PDF: 134
- XML: 54
- Total: 406
- Supplement: 53
- BibTeX: 39
- EndNote: 36
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Ehsan Erfani
Robert Wood
Knut von Salzen
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(11672 KB) - Metadata XML
-
Supplement
(1788 KB) - BibTeX
- EndNote
- Final revised paper
Lucas et al. present an interesting study in which they use Langevin particle modelling applied to realistic LES outputs to investigate the importance of the emission spreading rate of sprayers (ships) for Twomey forcing. They further examine the influence of variability in aerosol properties, including the number of sprayers and their motion type, aerosol decay timescales and age, concentration bins, and background aerosol concentrations, as well as meteorological conditions. Their results show that the velocity and type of sprayer motion have little influence on Twomey forcing. Assuming that cloud adjustments are relatively small, they further demonstrate that the effect of the spreading rate on Twomey forcing is limited, with the forcing being strongly controlled by aerosol lifetime. Their findings suggest that the common assumption in GCMs of an infinitely fast spreading rate could result in a 10–200% overestimation of Twomey forcing. Overall, I found the manuscript to be well written, and the methodology to be sound. I have only a few minor comments that can be addressed easily by the authors.
Minor comments:
1. Line 43: Goren et al. (2025) also attributed the co-variability to two factors: (1) large-scale meteorological conditions, which influence cloud properties such as cloud liquid water path (LWP) and cloud droplet number concentration (Nd), causing these variables to covary simultaneously and inversely; and (2) microphysical processes associated with cloud development, during which an increase in LWP is generally accompanied by a decrease in Nd.
Goren et al. (2025): Co-variability drives the inverted-V sensitivity between liquid water path and droplet concentrations, Atmos. Chem. Phys., 25, 3413–3423. https://doi.org/10.5194/acp-25-3413-2025
2. Line 45: By constraining cloud morphology using MODIS observations, Goren et al. (2026) reported contrasting results, finding that LWP adjustments nearly offset the Twomey forcing.
Goren et al (2026): Beyond discrete stratocumulus regimes: a ternary continuum of morphology reveals within-regime variability in cloud susceptibilities, Atmos. Chem. Phys., 26, 7193–7206. https://doi.org/10.5194/acp-26-7193-2026
3. Line 75–76: Are there any references demonstrating that these 17 LES simulations are in close agreement with the observations. Or the term “observations” is used loosely here to represent general shallow marine clouds. Please clarify.
4. Line 76: The abbreviation LPM appears for the first time here without being defined.
5. Line 118: Could the authors also describe how Nd is derived from the CERES measurements? In addition, how is Na computed from the MERRA-2 data? A brief description or an appropriate reference would be sufficient.
6. Line 263: This assumption is inconsistent with Equation (2), where only 80% of aerosol particles, rather than 100%, are assumed to activate as cloud droplets.
7. Line 272: Please define NAc
8. Figure 3: Please define what the top panel in each panel group represents in the figure caption.
9. Line 391 and Figure 10 (also in other instances): Is the twomey forcing calculated for these simulations “global”?
10. Section 4: Could the authors also comment on the role of cloud morphology and cell size on the spreading rate and Twomey forcing relationship?