We thank the three reviewers for their assessment. We have now responded to all the suggestions and have marked our changes to the paper in blue. > Reviewer #1 > 1. Abstract, l. 31: "...we show that both sources of fluctuations > on their own give an accurate representation of fluctuations.". > I am not sure if I understand this sentence. In the previous sentence, we wrote that the "algorithm incorporates fluctuations in two distinct ways: through the random spatial distribution of superdroplets and through the Monte Carlo collision algorithm involved". This refers to what is later called approaches II and III. To clarify the connection with "...show that both sources..." in the sentence above, we have now replaced "both sources of fluctuations" simply by "both", so it is clear that we refer to the previous sentence. > 2. L. 97/98: "The LDM assumes that the lucky droplet is large compared > to the background droplets, so that the radius of those smaller droplets > can be neglected in the geometrical collision cross section and velocities > of colliding droplets". Is this really true? Where is that assumption > made? We have now added "In its simples form," and refer to Eqs.~(3) and (4) of Kostinsky & Shaw; see the penultimate paragraph of the introduction. It is here where the LDM is stated in its simplest form." See the penultimate paragraph of Sect.1. > 3. Eq. 3. Lambda is not defined until eq. 4. We have now defined it directly after Eq.(3). > 4. L. 131: xi_i and xi_j are not defined until l. 142. We agree and have now moved the definition from below to just after Eq.(4) by writing "... of the multiplicities xi_i and xi_j of superdroplets ..." > 5. The conclusion section 6 is relatively short. Is that all what we > learned from the simulations described in the paper? Perhaps section 5 > and 6 can be merged? The conclusion has three paragraphs and we answer there the essential questions posed in the beginning of the paper. Our current discussion section reflects on our results in view of the paper by Dziekan and Pawlowska, which wasn't done until here, so it would not be suitable material for the conclusions. We have therefore decided to keep the current arrangement as it is. ----------------------------------------------------------------------------- Reviewer # 2 > 1) [request] P. 8, ll. 156--161, "It is then assumed that, ..." From > what you explained in your previous reply, I now understand that when > superdroplets with multiplicities \xi_i=0.5 and \xi_j=2.0 collide and > coalesce, you will remove the superdroplet i. When doing so, how do you > update j and guarantee the mass and momentum conservation? I think this > is an important detail that has to be explained clearly. Assume i is a > lucky superdroplet and j is a background superdroplet, and if you just > remove i from the system without changing the attributes of j, this > results in an unphysical partial destruction of the lucky droplet. As > pointed out below, this situation would be indeed happening in the (n_0, > n_luck0) case. Please elaborate. The text below Eq.(7) refers to the case xi_i=xi_j, stated in the beginning of the previous sentence. To clarify this better, we have connected the two sentences now by writing "and it is then assumed...". The removing does not apply to the reviewer's case "\xi_i=0.5 and \xi_j=2.0". Also, we have stated again that the mass and momentum are conserved by adding ", still conserving mass and momentum." > 2) [request] P. 21, l. 414, "However, this shortcoming may also be a > consequence of our choice of having used only 256 superdroplets,..." and > Fig. 8. Now you admit for the first time that in your original setup > described at the end of Sec.2, almost 50% of the background droplets > will be consumed when the lucky droplet grows to 50um. The result of the > original setup is shown in Figs. 6 and 7 already, but I strongly believe > it is informative to the readers to show this also in Fig.8(a), together > with the other cases for sensitivity test. If I understand correctly, > the original setup corresponds to (n_0, 10n_luck0), but it is not shown > in Fig.8a. We have now added the results from Fig.6 (same as in Fig.7) to Fig.8(a) and (b); see also the blue text in the caption. > If I understand correctly, in the (n_0, n_luck0) case denoted by the > red points in Fig. 8(a), the multiplicity of the lucky superdroplet > is \xi_luck(t_0)=0.1 (less than one!). Is this correct? If so, if we > apply your superdroplet removal rule, it seems to me that the lucky > superdroplet will be removed i the first collision-coalescence event, > but obviously this is not what you intended. Please clarify. As we just emphasized, droplets are only removed in the case described in Fig.1(c), when the two multiplicities are equal. > Please put the labels (a), (b), and (c), to Fig. 8. We have now added them. > Please specify the unit m^-3 of n_0 and n_luck0 placed in the upper > right of Fig. 8. We have now specified the unit. ----------------------------------------------------------------------------- Reviewer #3: Dear Authors, > - whenever you use the word "explicitly" (e.g. abstract), ensure that > it is clear what the opposite "implicit" approach would be We have now removed the word "explicit" in the abstract and only mention it in Section as "Approach II is an explicit method compared to other approaches listed in Table 3." below Eq.(17). > - p6: "limit the integration step through ... (Shima et al. 2009)" -> > wrong reference, Shima's formulation allows for long timesteps and p>1 We have now referred it to "Johansen et al. 2012; Li et al. 2017". > - p8: line 167: suggest clarifying "per time step" (should be per "sub > time step") and mentioning that the probability is up-scaled accordingly We didn't actually see anything about sub-time step or upscaling in the Shima+09 paper nor in the Johansen+12 paper and were worried to just adopt this without real knowledge. Since this discussion does not concern what we are doing, we feel that our current wording should be ok. > - p10: lines 224-225: this sentence does not convey any information > to the reader, likely candidate for removal (or is it referring to > the Appendix? - then refer to it) Yes, it does, so we have now referred it to appendix I. > - p11, eq. (10): what is n^{back} ? It should have read simply "n", and we have now changed it. >- p12: line 278: surplus "that" We have now removed it. >- p20: Fig 8 middle panel lacks legend entry for orange color We have now added the legend for the orange dots in panel (b). >- p25: Fig 12 caption mentions "thick black line", shouldn't it be "dashed black line"? We have now corrected it as "thick black dashed line." >- p29: zenodo url has the "doi.org" misplaced We have now corrected it as "https://doi.org/10.5281/zenodo.4742786" >- references: please add DOI identifiers We have now added all the DOI identifiers. >- references: fix capitalization in (multiple occurrences): > lagrangian -> Lagrangian; warsaw -> Warsaw; kuiper -> Kuiper; > neptunian -> Neptunian; earths -> Earths; scale-sdm -> SCALE-SDM; ii -> II, ... We have now fixed the capitalization.