Review of “A double-box model for aircraft exhaust plumes based on the MADE3 aerosol microphysics (MADE3 v4.0)” by Sharma et al.

Sharma et al. has developed a box model for investigation of the role of aerosol microphysics in the exhaust plume stage for aviation-induced aerosol number concentrations, comparing this to the case of assuming instantaneous dispersion as is done in more complex but coarse atmosphere and climate models. Aerosol-cloud interaction is one of the key uncertainties in assessments of aviation-induced climate impacts with estimates currently exhibiting a very wide range and strong dependence on underlying assumptions. Moreover, advanced understanding of plume processes, both chemical and microphysical, is needed. The study thus provides a timely and important contribution to the literature with potential for broader application. I have some comments/suggestions/questions, mostly for clarification, that should be considered before publication.

Broader comments:

• The study briefly mentions that the difference between aerosol number concentrations between the instantaneous dispersion and plume approaches can have “significant effects implications for the calculation of the climate effect of aviation aerosol on low clouds”. I would however encourage a somewhat more detailed discussion of what that effect might look like, referencing different assumptions in literature estimates and relating to the sensitivity of results shown here, e.g. in what direction this could push estimates of the ERFaci, implications for studies exploring future aviation fuels or propulsion systems (e.g. given dependence on SO2 emissions, limitation that this study is initialized with contrail ice typical for jet fuel), potential changing geographic traffic pattern etc. This would strengthen the motivation and broader relevance of the study.
• There are quite a few experiments performed. While they are described in each sub-section of the Results, I think it could be helpful for the reader to rather have a separate section for experiments under Methods, including a table with the experiment names
• Some figures, e.g. 11, 13 are a bit difficult to read – is it possible to increase font size?
• The abstract, introduction, and conclusion sections are quite repetitive, more or less listing the same results. Consider shortening the introduction and/or sharpening the abstract?

Specific comments:

Line 1 and elsewhere: I would encourage different wording than “significant” when there is no statistical basis for the statements.

Line 318: does this initialization also differ between regions or is it fixed?

Line 335: it would be helpful with a bit more detail in this paper about what “typical background concentrations” mean in EMAC, for the different regions

Line 342: If I understand correctly (otherwise, please clarify), this means that there is a range of estimates of the correction factors for each region? Given the sensitivity of the results to different parameters, this would seem like valuable information to add to the percentage numbers given throughout the text. Moreover, does “multiple ensemble simulations” here correspond to the number of ensembles given in table 1? If so, the number of ensembles differ substantially, does that affect the results?

Line 366: “such as SO2” – what else? It does not seem that this study considers species other than sulfate and soot?

Line 406: what does “typical conditions” mean – annual means? Based on frequency or occurrence? Do the authors expect strong seasonal differences given the regional variation found? Please define/specify.

Line 449: Maybe a column could be added to table 2 given the initial background concentration in the different regions? (And temperature and RH differs in table 2 – these differences are included? Would these differences be expected to influence the results? Via sulfate production, but likely small effect…)

Line 551-554: It could maybe be good with some absolute numbers as well, not only percentage?

Line 618: “remarkable” – is this the right word here?

This paper presents a new model to improve estimates of how aviation emissions affect climate through aerosol interactions. Specifically, the authors develop a double-box aircraft plume model that simulates aerosol microphysics in a dispersing exhaust plume, including interactions with contrail ice particles. This approach contrasts with the instantaneous dispersion method used in most global climate models, which oversimplifies plume-scale processes. The study shows that accounting for plume-scale microphysics—especially contrail ice and sulfur chemistry—can significantly alter estimates of aerosol number concentrations due to aviation.

The paper addresses an important issue: the treatment of aviation aerosol emissions in global climate models. It clearly outlines the deficiencies of the “instantaneous dispersion” approach and motivates the need for a more physically realistic treatment of plume-scale processes. This is a well-executed and timely study that advances the modeling of aviation-induced aerosol impacts. Its strength lies in blending computational efficiency with a more realistic physical treatment of plume evolution. Some simplifications are inevitable but should be better contextualized. With modest improvements—particularly around validation, applicability, and limitations—the paper will be a solid contribution to the literature on aviation and climate. The writing is generally clear and the paper fits well into the scope of GMD. I recommend addressing following clarifications/improvements before the paper can be considered for publication.

Major:

1. No feedback to cloud microphysics or radiative forcing. The paper discusses climate relevance, especially for low-level clouds, but does not actually quantify CCN changes or radiative impacts. Even a simplified treatment or discussion of how the aerosol differences would translate to cloud effects would enhance the paper’s broader significance.
2. Scope of chemical processes. The model only considers H₂SO₄ formation from SO₂ and neglects other chemical pathways or species (e.g., organics, NOₓ, secondary organic aerosol formation). A clearer discussion of why this simplification is acceptable (or its implications) would help define the model’s applicability.
3. Interfacing with a global model. Since this model is designed for efficient use in global simulations, it would be helpful to briefly outline the strategy for integration into global models, and how the plume correction could be used as a parameterization.
4. Limitations of modal models regarding size distribution predictions. While the authors note that the MADE3 microphysics scheme has been evaluated and found to perform well in global-scale applications, I would encourage them to consider the limitations of applying modal aerosol models in plume-resolved simulations. The physical and chemical conditions in aircraft exhaust plumes—such as steep concentration gradients, rapid dilution, and intense nucleation—differ markedly from the large-scale, averaged environments for which these schemes were originally designed. Recent work by Fierce et al. (J. Aerosol Sci., 181, 106388, 2024) highlights significant discrepancies between the modal MAM4 scheme and particle-resolved size distributions from PartMC, particularly under plume-like conditions. This raises important questions about the suitability of modal representations for capturing the microphysical evolution of aerosol populations in near-field plume simulations. I recommend that the authors discuss these limitations more explicitly and consider how they might affect the accuracy of their results and conclusions.
5. Presentation of the governing equations. The presentation of the governing equations would benefit from greater clarity and completeness. Currently, mathematical expressions are scattered throughout the text, but they mostly appear to be discretized or implementation-specific snippets rather than a concise formulation of the core model equations. To aid reader understanding and improve transparency, I recommend that the authors clearly define the governing equations at the outset of the methods section, beginning with a clear statement of the state variables (e.g., aerosol number, mass, moments, gas-phase species) and their dependencies (e.g., time, space, size). A brief but coherent listing of the continuous (pre-discretization) equations governing aerosol dynamics, gas-phase chemistry, and aerosol–gas interactions would greatly improve the readability and reproducibility of the study. While this level of detail is sometimes omitted in our field, including it here would strengthen the paper’s clarity and accessibility, particularly for a paper published in GMD.
6. Comparison to measurements. While I recognize that this is a process-level plume model and not intended to reproduce specific observations, I think it is still important to address the question of model validation or evaluation. Ultimately, we would want to know whether applying this modeling framework leads to improved predictions—either of aerosol properties in the far field, or of emergent metrics like CCN or optical properties. I understand that direct comparison with measurements may not be feasible at this stage, but some discussion of how the model’s outputs could eventually be evaluated—e.g., through comparison with aircraft observations, or as part of a nested modeling framework—would strengthen the manuscript. Even a statement clarifying that this work is a foundational step toward such evaluation would help position the contribution more clearly.

Minor:

1. Equation 1: l.h.s (rate) and r.h.s. (concentration) are inconsistent.
2. L. 226: “only” appears twice.
3. L. 232: “gets” is very colloquial. Suggest “is”.
4. L. 250: “same size” -> same mode
5. L. 276: value of  s is unclear – should this be 0.004 s^-1?
6. Equation 8: C_i^SP(t) appears on the r.h.s. – is this correct?
7. L. 311: should be ice crystal number concentration. This happens quite often throughout the paper, please check throughout.
8. L. 418: do you mean: “in terms of total mass and number concentration, and number size distribution”
9. Figure axes labels are too small in Figure 11 and 13.