In their article, Clusius et al. present the Atmospherically Relevant Chemistry and Aerosol box model (ARCA) representing atmospheric gas-phase chemistry and submicron aerosol processes. Its processes representation is mainly based on earlier model implementations but distinguishes itself by an extensive GUI and its focus on ease of use for scientists that have little experience in aerosol and chemical box modelling. Although the current manuscript provides already a good overview on the model itself, further elaborations are necessary before publication in GMD.

General Comments:

1. At the beginning of the introduction, a very short overview on possible applications for aerosol and chemical box models is presented. However, the examples given are solely limited to the published work of the authors. The introduction would greatly benefit when discussing the ARCA model in the context of other aerosol and chemical box models in a technical and historical context. Especially, focusing on models and publications outside the authors’ group.
2. As the authors clearly state (Sect. 1.1), the model in its current state neglects important processes for aerosols (e.g., aqueous-phase chemistry). However, the targeted user group is scientists that have little to no experience in aerosol and chemical modelling. This user group would greatly benefit from a section discussing the implications of the missing processes on the model’s results (e.g., before section 6).
3. The authors claim that the GUI design is one of the main aspects that distinguishes ARCA from other box models. However, very little information on the GUI design is presented and is mainly limited to a series of figures from ARCA’s GUI with very little information about it in the text (Sect. 3). Further elaboration on the GUI design is necessary in this section. This information should at least contain but is not limited to: (1) the GUI design principles, and (2) a workflow diagram and description of the GUI design/use including dependencies between the different tabs. At multiple locations in the text, the authors discuss possible errors. Since the targeted user group is expected to have little experience with box modelling, a proper error management is key to avoid misconfiguration followed by potentially wrong conclusion. An elaboration on the error message management should be added.
4. The verification of the model (Sect. 4) is poorly presented and e.g. in Sect. 4.3, it is limited to just presenting two plots. Some discussion of the results should be added.
5. Before the authors provide a revised version of the manuscript, they should perform an editorial check focusing on the consisted use of the Copernicus rules (e.g., use of figure within the text), correct numbering of sections, etc.

Specific comments:

L20: Stating that “any compatible chemistry scheme” is ambiguous since currently the model is limited to gas-phase chemistry.

L109: Is the user notified of this if the time step is selected too large by the user? Is there an upper limit of the time step length?

Figure 1: The definition of PSD is missing in the caption.

L271: The last sentence seems to be out of place. Shouldn’t this be specified in the introduction?

L285: Is the user made aware (e.g., in the GUI) that he is responsible to select the PSD size range? Is an error raised if he fails to do so? This could be used by the authors as an example for the error management (see general comment 3).

L397: What is your justification to select the same value for all compounds? Can the user change this value for single compounds? This would make it easier to tailor ARCA to individual chambers.

L432: The online user manual is nice. Adding a link to it here would be helpful.

L439: This statement is ambiguous, unspecific, and misleading. What distinguishes “good” and “better” science? I suggest rephrasing this statement focusing e.g. the avoidance of systematical errors when performing simulations or similar.

Section 5.3: In addition to the model code, please provide the modelling data of the simulations used in the verification section (e.g., using Zenodo).

Section 6: In Sect. 1.1, you state that organic liquid phase chemistry is not considered in ARCA but may be important (e.g., on the thermodynamics, formation of secondary organic aerosols). Is there a reason why you currently do not plan to include this chemistry in future model versions? In the case when using MCM, the coupling to CLEPS (Mouchel-Vallon et al., 2017) should be strait forward.

References:

Mouchel-Vallon, C., Deguillaume, L., Monod, A., Perroux, H., Rose, C., Ghigo, G., Long, Y., Leriche, M., Aumont, B., Patryl, L., Armand, P., and Chaumerliac, N.: CLEPS 1.0: A new protocol for cloud aqueous phase oxidation of VOC mechanisms, Geosci. Model Dev., 10, 1339–1362, https://doi.org/10.5194/gmd-10-1339-2017, 2017.

This manuscript presents the ARCA model, a fairly sophisticated box

model with a graphical user interface that can be used to simulate

atmospheric chemical processes. There are now several of these tools

available but not all of them are aimed at researchers without a

modelling background. I think the effort to simplify the process of

setting up and running a complex chemical model is noteworthy and it

seems ARCA fits into this description. The paper is well written and

clearly outlines the inner workings of the model, referring the reader

to the extensive online documentation for more detailed information.

I think there are only a few minor points to clarify, otherwise I

recommend publication in GMD.

The idea of using different timesteps for different processes is

interesting. I am wondering if the authors can say how much

computational time is saved compared, for example, to the fixed

timestep mentioned on line 121. I imagine this is related to the

chemical conditions used, but it would be good to have a ballpark

idea. One things that is not clear to me is whether the possibility

of using different timesteps for different processes is a consequence

of the fact that the processes are run in a sequence (line 105). As a

side note to this point, it would be good if the authors can comment

on how the model ensures that mass is conserved when "moving" from one

process/module (e.g. gas phase chemistry) to the next in the sequence

(e.g. particle formation).

On line 94 it is said that all particles are considered to be liquid

droplets, but on line 98 it is said they are void of water. This

sounds like a contradiction.

Is is also not clear to me what is the difference between the aerosol

module mentioned in this paragraph (lines 94-99) and the ACDC

module. The following sentence (lines 100-101) implies they are both

used in the model, but the text on page 2 (lines 73-76) seems to

suggest that the model uses the ACDC only. The APC scheme is also

mentioned on page 2, but it is not clear whether it is part of the

ARCA model or not.

I think there should be a little more explanation about the wall loss

process described in section 2.7 (and also about the "gas-wall

partitioning" in Figure 1. What is it meant for? Indoor studies and/or

environmental chambers? Is it inactive in ambient studies?

line 81: delete "using"

line 99: instead of "restrictions" I suggest "limitations"

figure 1: I suggest changing "wall loss of particles" to "particle

losses" for consistency with table 1 and related text (or viceversa,

change the table).

line 226: can you clarify what is meant by "numerical diffusion"?

line 271: it is said here that "ARCA has slots for five ACDC

modules". Does it means that there are 5 aerosol bins, as default?

Related to this, does the ARCA GUI allow modification of the ACDC

settings or one would need to modify those separately, following the

ACDC user manual?

line 511: add link/reference to "UmanSysProp" and say what it is.

line 517: why do you say KPP is not necessary for ARCA? That seems to

contradict the statements in section 2.1 and 2.1.2. The VODE (or

DVODE? Is it the same thing?) solver is mentioned only once earlier in

the paper. Please clarify if it is VODE or KPP (or perhaps both?) that

are used to solve the differential equations system.

In Section 5.2. Are there any libraries or dependencies that need to

be installed besides the netcdf libraries? e.g Python graphic

libraries for the GUI and plotting?