Review of ICON-HAM-lite: simulating the Earth system with interactive aerosols at kilometer scales:

General Comments:

1) Does the paper address relevant scientific modelling questions within the scope of GMD? Does the paper present a model, advances in modelling science, or a modelling protocol that is suitable for addressing relevant scientific questions within the scope of EGU?

Yes.  A new modal aerosol model is developed for use in sub-10 km chemistry transport modeling.

2) Does the paper present novel concepts, ideas, tools, or data?

Yes.  HAM-lite is a novel adaptation of HAM.

3) Does the paper represent a sufficiently substantial advance in modelling science?

Yes.  The authors demonstrate the applicability of HAM-lite for high resolution modeling.

4) Are the methods and assumptions valid and clearly outlined?

Yes, but with a caveat.   The authors describe their model experiment sufficiently.   But more information should be given about the performance of HAM-lite compared to HAM.   

5) Are the results sufficient to support the interpretations and conclusions?

Yes.  The model output is evaluated comprehensively and clearly.

6) Is the description sufficiently complete and precise to allow their reproduction by fellow scientists (traceability of results)? In the case of model description papers, it should in theory be possible for an independent scientist to construct a model that, while not necessarily numerically identical, will produce scientifically equivalent results. Model development papers should be similarly reproducible. For MIP and benchmarking papers, it should be possible for the protocol to be precisely reproduced for an independent model. Descriptions of numerical advances should be precisely reproducible.

Yes.  The model experiment presented in the paper can be reproduced using the same model and inputs by other authors and can also be approximated by other models making similar parameterization choices and using similar emissions and surface data (SSTs and sea ice).

7) Do the authors give proper credit to related work and clearly indicate their own new/original contribution?

Yes.   All relevant input for this paper is cited and original contribution clearly delineated.

8) Does the title clearly reflect the contents of the paper? The model name and number should be included in papers that deal with only one model.

Yes and no.   The model version numbers are missing but the title properly describes the paper contents.

9) Does the abstract provide a concise and complete summary?

Yes.

10) Is the overall presentation well structured and clear?

Yes.  But more content needs to be included as outlined in the Specific Comments.

11) Is the language fluent and precise?

Yes.

12) Are mathematical formulae, symbols, abbreviations, and units correctly defined and used?

Yes.

13) Should any parts of the paper (text, formulae, figures, tables) be clarified, reduced, combined, or eliminated?

No.

15) Are the number and quality of references appropriate?

Yes.

15) Is the amount and quality of supplementary material appropriate? For model description papers, authors are strongly encouraged to submit supplementary material containing the model code and a user manual. For development, technical, and benchmarking papers, the submission of code to perform calculations described in the text is strongly encouraged.

N/A.

Specific Comments:

1) This model lacks secondary aerosol formation (sulfate and SOA) and this should be explicitly mentioned in the model description. Lack of secondary aerosol affects the optical depth (Figure 6).  The authors attribute the deviation from MODIS observations to primary emissions issues on page 13.   This is only partially valid.  Lack of SOA in the model has a large impact on the carbonaceous aerosol simulated.   AeroCom models compared in Table 5 included SOA schemes.   SOA accounts for more than half of organic aerosol globally.   There is no mention of secondary aerosol anywhere in the manuscript.   The paper needs to be revised to outline this model limitation and its impact on comparison with observations.  

2) The resolution of 5 km is not sufficient to resolve convective updrafts and downdrafts which for deep convection have a diameter under 3 km over land and under 1 km over the oceans. Shallow convection is even smaller in scale.    Instead of trying to run with a compromise resolution globally, a better approach would have been to use a high-resolution regional domain driven by boundary conditions form a coarser resolution global run.    The authors should include more discussion about their choice of the 5 km resolution. Using the term “kilometer scale” to describe a 5 km resolution is not valid even though this has become a widely used term.    This resolution does not have the dynamical process resolving power of 2 km let alone 1 km.

• The 5 km resolution is in a transition or gray zone between the need for parameterized deep convection at 10 km and relatively reasonably resolved deep convection over land at 2 km. It is possible to apply a scale-ware deep convection scheme at transition zone resolutions (https://doi.org/10.1029/2021MS002696) since resolved convection is inadequate.  In terms of the smaller scale shallow convection, which is important for tracer transport, the 5 km resolution requires use of a parameterization scheme (https://doi.org/10.1175/MWR-D-19-0030.1).  
• The 5 km resolution is not good enough to reject using a 10 km resolution and assuming hydrostatic conditions with deep and shallow convective parameterizations.   This would have the benefit of being able to run with a 5-minute timestep instead of 40 seconds.   More resources could be expended on aerosol process representation.    In spite of progress in computer speed we are still not at the stage of running global cloud resolving models even with highly simplified chemistry and aerosol processes.

3) There is no evaluation of the HAM-lite model against the original HAM cited.  Comparison of the two schemes for test cases on small domains would give insight into the biases introduced by the simplifications in HAM-lite.   It is not particularly clear from the paper how HAM-lite parameters were tuned.   This subject is work for another paper, but the authors should include more discussion about how HAM-lite compares to HAM in terms of predicted aerosol distributions and how HAM-Lite parameters were selected. 

Technical Comments:

L104: Replace reference with Tegen et al. (2002): DOI: 10.1029/2001JD000963

L183: Correct spelling of asymmetry. 

The manuscript “ICON-HAM-lite 1.0: simulating the Earth system with interactive aerosols at kilometer scales” by Weiss et al. presents a new simplified aerosol scheme that allows to run global interactive aerosol simulations at km-scale grid spacings. The authors explain the implementation of the scheme and its interaction with the other physics schemes and show results from a yearlong simulation at 5 km horizontal grid spacing. The results show that the model can capture aerosol properties similarly well as state-of-the-art models and compares reasonable well compared to observations although systematic biases seem to exist. The emergence of mesoscale details such as dust storms on the leading edge of cold pools is particularly encouraging since such features have not been able to simulate at global scales so far. The paper is well written, and the images are of high quality. My only major comment is that adding some kind of comparison to the original ICON HAM model and a discussion about the differences would be important. The main question to answer here would be what the HAM-lite model cannot do what the more complex HAM model can and how does this affect the simulations and the ability to answer science questions. More details on this topic and a few more specific comments can be found below.

General Comments:

1. You mention that it would be too expensive to run the two-moment module HAM with 5 km grid spacing but I was hoping to see an intercomparison between the original HAM and HAM-lite even if it is only at 10 km or for a short period of time. Such a comparison would shed some light on the differences and potential implications of the simpler scheme and might also help to improve HAM-light. I understand that adding such a comparison would be a lot of work but would also bring a lot of benefits. If you decide to not work on this, I recommend to at least discuss potential differences and work on a comparison in the near future.
2. Related to the above comment; Could you add a discussion on what the systematic impacts of using a single moment aerosol representation are? What processes would be better captured using a two-moment representation and where should users be careful when using the one moment output for process understanding?

Specific comments:

L19: Do CMIP6 type models truly have complex microphysics compared to e.g., modern weather forecasting models. My impression was that they usually use single moment schemes without graupel/hail. Additionally, running deep convection schemes that produce precipitation complicates things.

L84-5: Currently this reads like there are three schemes for microphysics, radiation, and turbulence each. You could write: "There are three parameterization schemes, one for cloud microphysics, one for radiation, and one for turbulence..."

L109: "To forward emissions to modes,..." what is meant here? Please reformulate.

L259-60: I do not see anything special at 500 hPa in the dust concentration profiles. It seems as if the peak of emissions is related to the height of the source region (e.g. dust emissions over the Tibetan Plateau should be around 600 hPa). Plotting the profiles regarding height above surface should help to reduce this effect.

L269-70: I appreciate that you mention satellite observation uncertainties. Could you add a sentence on how those might affect this comparison?

L296: "The vertical velocities highlight diverging cold pool edges that lift air from convective downdrafts." This should probably be something like: ... diverging cold pools that originate from convective downdrafts and mesoscale circulation that lifts air at the gust front.

L299-300: It is not only the cold pools that are not captured but also the mesoscale circulation (e.g., mesocyclones, rear-inflow jets in MCSs...). One study that discusses this is:

Prein, A.F., 2023. Thunderstorm straight line winds intensify with climate change. Nature Climate Change, 13(12), pp.1353-1359.