the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Mar 2026
Sensitivity of Marine Cloud Brightening over the Great Barrier Reef to Spatial Variability in Aerosol Forcing: A Case Study using convection-permitting model
Abstract. The Great Barrier Reef (GBR) is increasingly threatened by mass thermal coral bleaching events under climate change. Marine cloud brightening (MCB) has been proposed as a potential adaptation strategy to reduce thermal stress by enhancing cloud reflectivity through aerosol injection. This study evaluates the sensitivity of cloud–aerosol interactions to aerosol emission intensity and spatial configuration over the GBR using convection-permitting Weather Research and Forecasting (WRF) model simulations.
A control simulation representing a non- to weakly-precipitating shallow trade-cumulus regime is compared with three MCB sensitivity experiments: a densely distributed (20 km apart), moderate-intensity emission scenario (EXP20), a sparsely distributed (100 km spacing), high-intensity scenario (EXP100), and an intermediate configuration (EXP40). Results show that enhanced aerosol emissions substantially increase near-surface aerosol concentrations, with dispersion strongly governed by source spacing and prevailing trade winds. The EXP20 configuration produces more homogeneous and widespread aerosol enhancements, whereas EXP100 generates localized peaks that are rapidly scavenged, resulting in smaller domain-mean increases despite identical total emissions.
Over a 24-hour period, domain-averaged cloud droplet number concentration (CDNC), optical depth, and cloud albedo exhibit strong sensitivity to aerosol loading, while cloud water path (CWP) and cloud fraction show limited responses. These findings indicate a dominant Twomey effect in this cloud regime, with only weak evidence of the Albrecht effect. Nonlinear CWP responses are noted under varying conditions of mid-level humidity, wind shear, and lower-tropospheric stability. Overall, the results highlight the importance of aerosol source configuration and background atmospheric state in shaping MCB effectiveness over the GBR.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Atmospheric Chemistry and Physics.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.- Preprint
(2073 KB) - Metadata XML
-
Supplement
(368 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-1251', Anonymous Referee #1, 20 Apr 2026
-
RC2: 'Comment on egusphere-2026-1251', Anonymous Referee #2, 27 Apr 2026
General comments:
This paper reports on simulations, using WRF, of a case study in the Australian GBR region, of how cloud microphysical and macrophysical properties are affected by marine cloud brightening sea salt aerosol injections. Three injection cases are simulated, all with the same total mass of injected sea salt, but with different distributions of point sources (EXP20: 75 sources; EXP40: 12 sources; EXP100: 3 sources). The EXP20 case produces the largest cloud albedo changes, and this is shown to be mostly via the Twomey effect. The paper further breaks down the results from the EXP20 study to quantify differences in cloud brightening driven by three metrics for meteorological controls: above-cloud humidity, wind sheer below vs above cloud, and boundary layer estimated inversion strength (EIS).
In particular, the contribution of examining the effect on MCB efficacy of spray system spatial distribution is a valuable addition to the literature, and overall the analysis is technically sound. The paper is generally well-written and well structured, and the figures clear. However, I think the paper includes statements that I think are either too strong or unsubstantiated by the presented analysis, and in a few places where wording needs clarification.
Specific comments:
lines 21-24: It’s not clear what’s meant by “intensity”. Is this just referring to the number of emission points? Or the rate of emission at each emission point? It’s not clear from the text here that all simulations have the same total emissions, just distributed over a different number of emission points, as described in the main text. This text should be edited for better clarity.
lines 134-136: What size aerosol are emitted? This is important, given the strong dependence of aerosol activation on aerosol size.
line 193-197: In reference to the comparison of the simulated (panels a and d) vs observed (panels b-c and e-f) cloud fields, shown in Figure 4: I’m struggling to agree with the authors that the simulations capture the cloud field “reasonably well”. The cloud fraction in the study area is quite different (given later as ~7% in the simulations but what looks more like 50-60% in from Himawari obs), as is the morphology of the cloud fields (very small, broken, evenly distributed Cu in the simulation, vs broader and more variable areas of cloud cover in the obs).
This leads me to think that the boundary layer thermodynamics might be different in the model than in reality. Figure 3 shows good what looks like very agreement between the simulated and observed temperature, dew point and wind profiles, but for these clouds what will most matter is what’s happening in the lowest few km of the atmosphere. In Figure 3a, the simulated and actual winds below ~800hPa are actually pretty different; in Figure 3b it’s hard to see what the sounding winds are doing.
-> I’d suggest adding two additional panels to this figure, reproducing the current panels a) and b) but zoomed in on, e.g., 700mb to the surface.
-> Further, I think this statement needs to be softened. The truth is that it is really challenging to reproduce the specifics of a given cloud field, even with a perfectly initialized simulation. This can be acknowledged in the context of not overstating how well the model is reproducing reality.
Figure 5 vs Figure 6: Comparing these two figures it appears the model is only activating about 15% of the aerosol to CCN – an activation rate that seems very low. This could be due to the resolution of the model (1km in the inner domain; 5km in the outer domain), as the lower resolution could be leading to a low bias in updraft rates. Ultimately this will produce a bias in the mass of aerosol that needs to be emitted to achieve a given forcing. Perhaps in the discussion, the simulated aerosol activation rate should be put in the context of expected values for typical low marine stratocumulus and this potential source of bias acknowledged.
Figure 6 a-c and Figure 7a: Why show the *surface* aerosol concentration, rather than the concentration at cloud base? The variability in concentration at the surface isn’t what matters; it’s the variation in concentration at cloud base that will affect CDNC.
(Note that I’m asking this based on the assumption that “surface” means the value in the model’s lowest layer, but maybe that’s incorrect? If you’re going to stick to showing ‘surface’ concentrations it would be good to at least define what you mean by ‘surface’.)
lines 241-244: The explanation given here of the precip scavenging efficiency at higher aerosol concentrations driving the difference in domain-mean concentrations across the different cases intuitively makes sense -- except that it is noted earlier that these are “non- to weakly-precipitating” clouds. Later, Figure S3 shows that the clouds were indeed only precipitating for a few hours in the middle of the simulation, and at very low rates (<0.01mm/hr). Can the simulated rain rate be used to quantify (even approximately) how much precip scavenging could be accounting for the differences in domain-mean aerosol concentrations across the different cases? And does this align with the simulated differences in concentrations? I think such an analysis should be included, especially given it affects the first main conclusion of the paper (lines 429-430)
More generally, is there a reason for ruling out other causes of the differences? e.g. Could coagulation or other processes be driving some of the difference? (This relates to the question of the injected aerosol size).
If a difference in precip scavenging is truly the main reason why the domain-mean aerosol concentrations differ, then the results shown here are very specific to the precip rates in this case, which brings into question whether the analysis of a single case allows one to reach a conclusion about the most efficient deployment configuration in general.
lines 321-323: I have to object to the assertion that the cloud albedo exhibited “pronounced increases” in the injection cases relative to the CTRL case. The increases are well within the range of variability in the CTRL case (Figure 8e); in Figure 7e, only the EXP20 case has a cloud albedo that is notably and consistently higher with aerosol injection than without.
lines 431-432:
This is an incomplete sentence.
Also, again I think that other than (marginally) in the EXP20 case, the cloud albedo increases were modest to (EXP100) negligible.
line 441, Conclusions: “The findings confirm that radiative forcing can be meaningfully enhanced under suitable conditions through spatially coherent aerosol seeding. As such, strategic deployment of MCB during favorable weather regimes holds promise as a targeted intervention to mitigate extreme heat exposure over sensitive marine ecosystems like the GBR”.
I feel strongly that these statements cannot be made. No radiative forcing calculations are shown. The paper later asserts that the cloud albedo change from EXP20 vs CNTRL could produce a forcing “potentially on the order of several tens of W m⁻²” but I don’t see how this is the case. In EXP20, the cloud albedo increases to 0.51 from 0.48. If there was 100% cloud cover this would, ballpark, result in an increase in reflected solar flux of ~10 W/m2 (downwelling 340 W/m2*(0.51-0.48)=10.2 W/m2). But cloud cover isn’t 100% -- it’s, in this case, 7%.
Further: Being able to say forcing is *meaningfully* enhanced, and that MCB could actually mitigate extreme heat exposure in the GBR, would require showing that the resulting forcing would be sufficient to produce ocean cooling.
All of this is well beyond the scope of the paper – but so is the assertion.
Unless the authors want to add this analysis, they can only assert what the paper has actually shown, which is articulated nicely on lines 446-495, and is of sufficient scientific value to stand on its own.
Technical comments:
- lines 68-69: A classic paper that should probably also be cited here is the Stevens and Feingold (2009; https://doi.org/10.1038/nature08281) paper that’s also cited in another context later in the paper.
- line 184: skilfully -> skillfully
- line 184-185: “The control simulation skilfully simulated the evolution of the synoptic-scale mean surface level pressure and surface winds (not shown).” The evidence showing this should be included in the Supplemental data/file.
- Figure 5: It appears that none of the CDNC values in any of the panels exceed 150/cm3, so why not set the colorbar range from 0-150/cm3 (or maybe even 0-125/cm3)?
- line 231: “In the contrast, …” -> “In contrast, the …”
- line 233: “exhibited” -> “exhibits” (for consistency w/ present tense used elsewhere)
- line 435: “mid-level” is a bit ambiguous; I’d say “above-cloud”
- line 524: this is the first use of LWP; for consistency with earlier usage, suggest changing to CWP
- Figure 9: The caption says that “colors with black dots indicate *significant* differences”. Is this correct? I suspect that this should read *insignificant* differences, given that the back dots correspond to when there is near-zero delta-density.
Citation: https://doi.org/10.5194/egusphere-2026-1251-RC2 -
RC3: 'Comment on egusphere-2026-1251', Michael Diamond, 04 May 2026
In their manuscript, the authors run a convection-permitting simulation of MCB over the Great Barrier Reef and show that a dense network of sprayers using moderate emission rates is substantially more effective at dispersing aerosol and brightening clouds than a sparse network of high-emission sprayers. The core idea is important and worthy of publication. Unfortunately, much of the supporting commentary requires major revisions before I can recommend the manuscript be published.
Major comments:
A) Interpretation of aerosol “scavenging”: As the authors show elsewhere, precipitation (and therefore wet deposition) is not particularly high in their simulations. Increasing aerosol would also tend to decrease precipitation. Denser aerosol plumes therefore would not be expected to result in greater wet scavenging; this explanation for the differences between EXP20 and EXP100 is implausible. More likely, the aerosols are coagulating more in the sparser-but-denser plumes, reducing number but not necessarily mass. An easy way to check this would be to look at the total or average mass concentration of aerosol between the experiments in addition to number concentration.
B) Overinterpretation of aerosol-cloud interaction results: The novelty of the current work lies in the exploration of the denser versus sparser sprayer setup, not in exploring ACI under shallow cumulus conditions. The general finding that adjustments to the Twomey effect would tend to promote cloudiness under moist and precipitating conditions and decrease cloudiness under dry and non-precipitating conditions is very well established at this point. The results and discussion should be rewritten to better highlight which areas of the study are a novel contribution versus which are reinforcing the preexisting consensus. The abstract already does a nice job of this!
C) Analysis of cloud and radiation variables: A number of issues arise with the treatment of cloud properties, including an inappropriate means of calculating cloud fraction. See specific comments below. I would also encourage the authors to consider evaluating cloud radiative effect (CRE; all sky minus clear sky net fluxes) directly.
Specific comments:
- Title: “A Case Study using a Convection-Permitting Model”
- Line 46: The most current review of MCB science is: Feingold, G., Ghate, V. P., Russell, L. M., Blossey, P., Cantrell, W., Christensen, M. W., Diamond, M. S., Gettelman, A., Glassmeier, F., Gryspeerdt, E., Haywood, J., Hoffmann, F., Kaul, C. M., Lebsock, M., McComiskey, A. C., McCoy, D. T., Ming, Y., Mülmenstädt, J., Possner, A., Prabhakaran, P., Quinn, P. K., Schmidt, K. S., Shaw, R. A., Singer, C. E., Sorooshian, A., Toll, V., Wan, J. S., Wood, R., Yang, F., Zhang, J., and Zheng, X.: Physical science research needed to evaluate the viability and risks of marine cloud brightening, Science Advances, 10, eadi8594, doi:10.1126/sciadv.adi8594, 2024.
- Line 52: As the authors note elsewhere, secondary indirect effects/cloud adjustments can also decrease cloudiness by increasing dry air entrainment
- Section 2.1: Please clarify if the nesting is one-way or two-way
- Line 195: Can the authors elaborate on the “key features of maritime shallow clouds, visible in the true-color imagery” here? In my read of Figure 4, the agreement is not at all obvious
- Lines 241-242: Precipitation scavenging efficiency increasing nonlinearly with aerosol concentration is not what I would expect… presumably this is referring to scavenging by impaction of aerosol on raindrops? Nucleation scavenging should be a much larger sink and will decrease in efficiency with increasing aerosol.
- Line 242: What is meant by “self-scavenging” here? Dry scavenging? Are the aerosol truly being lost (in terms of mass)? This sounds more like coagulation to me.
- Figure 6: What are the “-C” and “-U” after EXP40 and EXP100, respectively?
- Line 274: It does not seem like you have shallow stratocumulus cloud regimes here… cumulus?
- Lines 296-297: This is an inappropriate way to define cloud fraction, if I’m correctly interpretating this as dealing with 3D outputs, and results in values that seem wrong on their face (e.g., Figure S2 definitely shows cloud fraction is much higher than 7%!). The fraction of columns with cloud optical thickness or liquid water path about a given threshold (e.g., COT > 1 or COT > 3) would be a better choice here.
- Lines 313-315: I’m surprised COT needs to be estimated instead of output directly by WRF… it probably doesn’t matter, but you do have the information required to calculate COT explicitly
- Lines 329-330: Why not calculate CRE and show this?
- Figure 8: Something seems off in this figure. For example, in (f), the EXP100U
- Figure 8: Why are EXP40 and EXP100 given U’s in their names?
- Figure 9 and related discussion: This figure does not show your results very clearly, at least in my read. Figure 10 is much easier to interpret and there doesn’t seem to be anything Figure 9 is doing that Figure 10 doesn’t already show more clearly. Consider removing Fig. 9; perhaps you can add a third bin to Fig. 10 if you wanted to show finer-grained variability?
- Section 3.4: I would hesitate to use words like “marked” and “clear” here, these results really are not that dramatic… which makes sense, given elsewhere you discuss how the aerosol is not changing the cloud macrophysics much!
- Lines 438-440: I don’t disagree here, but the results in this paper alone suggest the meteorology/cloud adjustments really don’t matter much here…
- Line 471: “Our work extends this understanding to shallow maritime cumulus” is greatly overstated. There is an extensive literature at this point on cloud adjustments outside the core stratocumulus regimes. Chen et al. (2024) is one recent, prominent example that has essentially identical findings about cloud adjustments as those presented here: Chen, Y., Haywood, J., Wang, Y., Malavelle, F., Jordan, G., Peace, A., Partridge, D. G., Cho, N., Oreopoulos, L., Grosvenor, D., Field, P., Allan, R. P., and Lohmann, U.: Substantial cooling effect from aerosol-induced increase in tropical marine cloud cover, Nature Geoscience, 17, 404-410, 10.1038/s41561-024-01427-z, 2024.
- Lines 478-479: My read of the state of the ACI literature is that there’s a general consensus that LWP adjustments on net are relatively small and offset a fraction of the Twomey effect but cloud fraction adjustments are potentially quite large, maybe even rivaling or surpassing Twomey in importance. The revised Rosenfield et al. (2019) paper does show limited LWP changes, but what they originally attributed to LWP are actually cloud fraction increases (at least in their revised analysis).
- Line 529: Do the authors consider local MCB to be “geoengineering”? I don’t necessarily disagree, but the “intervention” phrasing used in the introduction is perhaps more accurate.
- Data availability statement: Where can readers find the model outputs needed to reproduce the analyses in the text?
- Supplemental figures: Consider bringing these into the main text
Citation: https://doi.org/10.5194/egusphere-2026-1251-RC3
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 862 | 324 | 52 | 1,238 | 143 | 75 | 81 |
- HTML: 862
- PDF: 324
- XML: 52
- Total: 1,238
- Supplement: 143
- BibTeX: 75
- EndNote: 81
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
See the attachment.