This manuscript is a very elaborate description of an active fogwater collector, which is employed at the Puy de Dôme site in France. The collector is precisely described, its internal flow is modelled with a CFD fluid dynamic model, the collector and its performance are tested in the field during side-by-side comparisons with two other fog collectors. Results are described in great detail, good performance and good agreement with the other 2 active collectors is shown. This refers to the sampling efficiency (with respect to the liquid water content (LWC) of the fog or cloud, and to the results of inorganic ion analyses. The manuscript is well written, well supported by data (including the supplementary material) and deserves publication in AMT after the following 3 concerns have been addressed accordingly.

1. Results of 17 cloud events are presented. Is that number (17) the entire population of events probed? If yes, please state that prominently in the manuscript. Or, alternatively, are these 17 “beautiful” events that were selected from a larger pool of data. If so, authors are asked to communicate in the manuscript the arguments / parameters / thresholds they used to select these 17 events. Important: At some point in the manuscript, sampling dates (and times) of all 17 events need to be documented.
2. The title of the manuscript is awkward and needs revision. First, BOOGIE seems to be a nickname for the collector, it does not appear to be an acronym. However, the prominent mention of the word (capital letters in the title) suggests something extraordinary which is not present. Also, authors mention that the collector is newly designed (for example, in line 571), which is not really true since it has been employed for over 30 years. Further, the title of the manuscript promises more than the manuscript really offers. Of the second part of the title (…Biological, Organics, Oxidants, soluble Gases, inorganic Ions and metal Elements), this reviewer really only found inorganic ions. Biological analyses are mentioned, but results are not shown in a very limited way: ATP/ADP ratio data are shown in the supplementary, results briefly discussed in the main manuscript (lines 506 – 512). Oxidants and metal elements are not even mentioned in the results section. The latter would require special attention before the backdrop of the metal construction of the collector. Very few results are shown for formaldehyde and hydrogen peroxide, while it is not clear to this reviewer if they meant to represent organics or soluble gases.  All in all, the second part of the title needs to be removed.

1. Droplet size distribution (DSD) data are not shown or analyzed. It is suggested (lines 550, 551) that such data was not available. On the other hand, it seems that there is plenty of literature from that site showing and discussing DSD data. Authors are asked to clearly state in the manuscript that such data is not available in all 17 events. Alternatively, please make very strong arguments in the manuscript about the reason why such data was not employed in this manuscript. Alternatively, analyze such data to support arguments of collection efficiencies as a function of droplet sizes.

The authors present design information, along with select performance evaluation information, for a new cloud collector which they call BOOGIE. The collector is designed based on inertial plate impaction behind 3 rectangular jets and features an air sampling rate in the range of several existing collectors. For example, the air sample rate, as pointed out by the authors is slightly higher than the CASCC2 but remains well below the air sample rate of the original CASCC. The authors use CFD to estimate droplet collection performance and compare sample collection efficiency via comparison to optically measured cloud LWC and sample composition to measurements from two widely used samplers, the CWS and CASCC2.  Overall, the manuscript is well organized and the new collector design and evaluation should prove of interest to AMT readers.  There are, however, several items in need of attention before the manuscript should be considered for publication including one critical error that influences several analyses.

Major items:

1. I was surprised to see a jet impactor operating using a rotary bladed fan since jet impactors often have a relatively high pressure drop and the performance of these fans can strongly decrease at increasing pressure drop. It appears that this issue is mitigated by the relatively high drop size cut of the collector (10 um). The manuscript would be improved by a discussion about the pressure drop generated in the collectors and how this relates to the ability to collect small droplets.  Why was a 10 um size cut targeted rather than a more conventional value around 5 um? Was this a design choice or just how things turned out?
2. It is strange that the authors often present results in terms of their relationship to the exit velocity. Exit velocities behind a rotating fan are difficult to measure.  Further, the relationship between flow entering the collector and flow exiting the collector depends on the pressure drop through the collector.
3. Lines 164-165: The authors’ statement here about the theoretical flows entering and exiting the collector violates the principle of conservation of mass. The entering and exiting volumetric flow rates should be the same when adjusted for pressure drop through the collector.  I think the theoretical ratio here is meant to refer instead to velocities.
4. I am somewhat surprised that some fraction of droplets impacted on the vertical plates are not blown off the sides of the impaction plates, pulled by the airflow around those plates. The collection efficiency results suggest this is not, however, a major issue.  Can the authors comment on what prevents this from happening? Does the CFD simulation suggest there is a quiescent stagnation zone near the plate surface that prevents the drops from being pulled toward the plate edges?
5. There are a number of issues about the use and description of the CASCC2 that need to be clarified in a revised manuscript:
   • The normal CASCC2 as described by Demoz et al. has a polycarbonate body, Teflon collection strands, and a Teflon collection trough. Was the CASCC2 borrowed for this work modified to metal construction?
   • The sampling performance of the CASCC2 depends on the fan used to pull the airflow. The analysis in Demoz et al. is based on a 115VAC Nidec-Torin TA700 fan operated at 60 Hz.  What fan was used in the CASCC2 operated in the current study? The fan performance will change even with the 220 V/50Hz version of the TA700 fan.  In particular, the flow rate will decrease with a lower frequency voltage supply. This should be considered in the performance analyses.
   • The CASCC2 is sometimes operated with a downward facing rain excluding inlet. Was that used here?
6. Lines 357-360: The authors make a critical error here in the calculation of the BOOGIE air sampling rate. They should not assume that the exit velocity matches the velocity through the impactor slots.  Rather, they should use conservation of mass and assume that the exit volumetric flow rate matches the total volumetric flow rate entering the collector, adjusting for the pressure drop through the collector. Their incorrect assumption of matching velocities introduces a critical error into the collector sampling flow rate that will bias many of their subsequent analyses.  This error needs to be fixed, sample flow rate corrected, and derived LWC and collection efficiency also corrected. This is a critical error and must be fixed!!
7. The same error discussed in item 6 appears to have been made for considering the CWS flow rate requiring similar correction and updating of results.
8. The authors should introduce and define the concepts of isokinetic and non-isokinetic sampling much earlier in the paper and address them when discussing collector performance. They should also be clearer in discussing when and how the BOOGIE collector was aligned with respect to the wind in the collection efficiency evaluation.
9. Figure 3. The authors should add drop sizes to the legend. In addition to the model results in Figure 3, the authors should present modeled collection efficiency (e.g., at 10 m/s) vs. drop size. This is how collector performance is typically shown.  Finally, the authors should use jet velocities rather than exit velocities as the independent variable in Figure 3.
10. The authors should more clearly point out that drop composition within a cloud has been experimentally shown to vary with drop size (e.g., Bator and Collett (1997) Cloud chemistry varies with drop size. J. Geophys. Res., 102 (D23), 28071-28078) and that differences in the composition of collected samples may therefore be expected if the lower cut size for the collector changes. They kind of dance around this concept but never explain it clearly.
11. Figure 5: I wonder if these comparisons would be better represented by scatter plots of concentrations measured in the collector pairs. As presented, it is unclear if larger differences might simply reflect measurements of species at very low concentrations.
12. The authors occasionally allude to BOOGIE being capable of supercooled cloud sampling. Has this been evaluated? Are the impaction surfaces easily removable to retrieve accumulated rime? Are there issues with the collector face or jet entrances riming and/or clogging? What about the fan? With drops up to 10 um transiting the fan I would think this could be an issue.

Minor items:

• The manuscript does not adequately describe turbulent conditions in the collector. As far as I can tell, the only evaluation of turbulence is through the CFD modeling and that is not well described. There are also ambiguous claims in the abstract and conclusions about “few turbulences have been observed…”.  Were there any “observations” of turbulence and, if so, what do these claims mean?
• Line 103: change “where and their designs” to “where their designs”
• Line 107: Change “Caltech University” to California Institute of Technology”
• Lines 112-113: Her the authors refer to “The sampler”. This is confusing as written.  I first thought this was referring to the BOOGIE sampler but I think it actually refers to the CWS.  Please rewrite this sentence to be clear.
• Lines 198-199: please specify here the optical instrument used to measure LWC (PVM-100).
• Line 223: replace “resistance” with “drag”
• Line 228: replace “strains” with “strands”
• The authors should define the drop effective radius and clearly point out that this is not the mean physical radius.
• Lines 396 and 397: Change “PWM” to “PVM.” Why can’t the PVM output be recorded at higher time resolution? It’s just a voltage.
• Line 435: The sentence here is garbled. Please rewrite.
• Line 531: Replace “in a radius” with “in radius”
• Line 548: Replace “would present” with “would likely present”
• Line 560: It would be worth pointing out that there is a version of the CASCC family specifically designed for supercooled cloud sampling: the Caltech Heated Rod Cloud Collector (CHRCC). This is discussed by Demoz et al. and other references.

Vaitilingom et al. present the design of a new bulk cloud water collector (BOOGIE) which has been tested and applied at the Puy de Dome station and will be used for studies of biological and chemical composition. The new collector is well-motivated by shortcomings of existing designs especially with regards to sampling rates and ease of operation for biological applications and its characteristics and comparisons with other designs are reported and discussed. I have a number of comments and suggestions the authors might want to consider and I recommend publication after these have been appropriately addressed.

Major comments

• The intercomparison of solute concentrations (inorganic ions, formaldehyde and H2O2) between different collector designs is at the core of the collector evaluation, because this is what these instruments are made for in the first place. Unfortunately, this part of the manuscript is a bit weak. Only two samples were taken in parallel with the CWS, which has been in operation for more than 20 years at the station, and three with the CASCC2, another popular design. Based on such low statistics, it is difficult to robustly judge on the comparability of cloud water solute concentrations. I wonder these are really the only available data to compare? Given that the first BOOGIE application took place 8 years ago already, I would have expected a larger database for comparison. In case there is more data, I would suggest to include it to the manuscript, which would much strengthen the intercomparison section.
• The intercomparison of ion concentrations is discussed in a bit of an optimistic way. Deviations of up to a factor of about 2 (Df < 0.5) are assessed as „good comparability“, but it remains unclear what this positive assessment is based on. Would systematic deviations in this range not pose a problem to the long-term trends of concentration data at the station? How do the deviations compare to earlier collector intercomparison exercises, for individual ions as well as overall?  In a few extreme cases, deviations up to factors of 3 and 6 were observed between BOOGIE and CASCC2, but these are hardly discussed. Instead, it is implied that CASSC2 concentrations „appear“ „slightly higher“ and only „at first glance“. I’d recommend to revise this section in a more detailed and more critical way. Overall, the results seem to call for more systematic intercomparisons with a larger number of samples.
• There are a number of inconsistencies for some quantitative results between abstract, main part and conclusions, which I will give below. In general, I’d suggest to critically check that correct and consistent information is given everywhere throughout the manuscript. Also, the wording and sentence structure would benefit from a critical check here and there.
• Fig. 4 and its discussion are a bit misleading. Based on the data points, I believe the authors have forced their regression through zero, which means their slope, intercept and coefficient of determination are incorrect. Forcing through zero should only be done for very good reasons. In this case, the intercept could actually be informative and should not be artificially removed. With a simple linear regression, I would expect a slightly negative intercept and a slope close to 1. Given that small unrecovered amounts of water from inner surfaces or even slight evaporation inside the collector seem plausible, I would not expect an intercept of 0. Also, the collection efficiency does not seem to decrease at higher LWC, based on just the data points. This seems to be an artifact from the applied regression through zero. Please carefully check and revise accordingly.
• L357ff: From the explanations, it seems outlet velocity (not inlet velocity) together with inlet surface area has been used to calculate the air volume flow rate. This would be wrong and would need to be corrected throughout the paper, including any discussions and conclusions that might change with a corrected air flow rate.

Further issues

• The title should be revised as it is uncommon to include two colons. Also, it becomes clear at second glance only, why oxidants and metal elements are mentioned. These are not measured and discussed in the manuscript, but seem to be part of the collector name. This could be made clearer, e.g. like this: “Design and evaluation of a new cloud water collector for the analysis of biological, …, and metal elements (BOOGIE). Actually, nowhere in the paper it is explained where the name BOOGIE originates from.
• Results of the collector intercomparison should be summarized in a more quantitative way in the abstract and conclusion sections. The purely qualitative and subjective statement of “comparable” concentrations is not very informative. To not lengthen the abstract further, the introduction on the importance of cloud water composition could be shortened.
• L28: Doe you mean “lightweight” or “simple” or “easy to operate” mobile sampler?
• L33 and L346: The mean collection rate is inconsistent between abstract and results section. Please check and correct.
• L60: What does HAP mean? Make sure to introduce all abbreviations at first occurrence.
• L103: delete first “and”
• L112: Was the sampler indeed obtained from Kruisz et al. or built according to their design?
• L256: Do you mean “More information on this analysis is given in …”?
• L306: When you say “classes of particles”, do you mean just different sizes? Or has anything else been changed for the different “classes”?
• L310-311: Incomprehensible sentence, please rephrase.
• Fig. 3 would be easier to understand, if droplet sizes were given directly in the legend, rather than arbitrary “class” numbers. Also, I’d suggest a more compact y-axis label without uncommon abbreviations, e.g. “number collection efficiency” and “mass collection efficiency”
• L315ff: It is not always clear which type of collection efficiency is discussed (number vs. mass). Please make sure to correct accordingly.
• L316: At velocities below 5 m/s, the collection efficiencies of large droplets are not always > 50%. Please check and revise sentence.
• L319: Does the given average collection efficiency not strongly depend on the droplet sizes which were simulated? If more larger sizes were simulated, the average collection efficiency would be different. Please clarify what can be learned from this calculated average efficiency?
• L324-328: This section is difficult to understand and partly redundant (50% cut-off at 10 microns). Please revise.
• L335: Why are the theoretical efficiencies assessed as “good”? Are there any comparisons with other data that would substantiate this judgment?
• L340ff: Please check for repeated, i.e. redundant information.
• L342: It would help the reader to give a brief summary of the main characteristics of the sampled events.
• L354: In the section before, the optimal velocity is given as 10, not 8 m/s. Please check and correct.
• L359: In L164, the total inlet surface is given as 11.088E-3, please check and correct.
• L373 and elsewhere: LWCmeas, not LWCmes
• L396: PVM-100
• L397: The acquisition rate of the PVM could easily be increased, but I doubt it is really a relevant factor here
• L398: Why is CLWCexp “intrinsically an estimate”? In the previous section it is explained as a value from measured velocities and geometries.
• L417/418: Harmonize between “radius” and “diameter”
• Table 1: Rain intensities during the events should be included as well. The information on possible rain drop contamination is given late in the discussion only, but is quite a relevant information.
• L438: Not sure if it is really advisable to use collected water volumes as a metric for cloud LWC.
• Fig. 5: What does the “analytical error” of 10% represent and how was it determined? In the Supplement, “accuracy” is mentioned, but precision or repeatability would be more appropriate metrics here. Is it really correct that this value is the same for minor ions with low concentrations and for major ions with usually larger peaks?
• Fig. 6 could go into the Supplement, as its discussion is brief and does not add much. Instead, Fig. S13 could be shown in the main manuscript as it presents a different aspect of the evaluation, i.e. comparability between two identical BOOGIEs. It might even deserve its own subsection.
• In Fig. S13, why are some of the aliquots much different from the other two of one and the same sample? Measurement precision from repeated IC analysis seems to be in the low percent range and can thus not explain these deviations. How much do these “inconsistent” aliquots statistically impact the between-collector comparison?
• L487: In the Supplement, it says “duplicate” analyses, not triplicate.
• L504/505: These comparisons should be made in a more detailed way, i.e. ion by ion and making sure the metric for “difference” is really identical.
• L505: Are these values consistent with what is discussed earlier?
• L506ff: The comparison of biological data might warrant its own subsection. Right now, it is a brief extension on the chemical comparisons, but a bit difficult to understand without more context.
• The conclusion section repeats much basic information and explanations from earlier sections. It could be improved by summarizing in a more compact way the core results and conclusions.
• L559: The results presented before do not seem to support the statement of high collection efficiency of the CASSC2, it was actually the lowest one among the compared designs.
• L560: How is the applied CASSC2 model not affected by rain? The data of the cloud event with slight rain do not seem to support this statement.