Review of “Characterization of a self-sustained, water-based condensation particle counter for aircraft cruising pressure level operation” by Weber et al.

The present manuscript deals with the performance of a new commercial condensation particle counter aimed for the automated operation onboard passenger aircraft. In a laboratory simulation of the flight conditions, the performance was characterized. The thorough characterization of such an instrument for standard operational use is an important task, and it may serve as reference for the deployment of this instrument at other locations. The manuscript shows the results of these calibration measurements and discusses deviates as function of thermodynamic conditions. It generally shows the suitability of the instrument for the intended use, onboard the IAGOS container.

The manuscript is suitable for the publication in AMT, but requires some clarification and corrections. A grammar check for punctuation is suggested.

Remarks/Questions

Abstract: It should be mentioned in the abstract that the instrument was modified after this investigation. Also, maybe a recognizable instrument version of MAGIC should be given to avoid misunderstandings of the applicability.

Figure 1: Many arrows are not straight, which is inadequate for a scheme. What is the difference between the line from the top labeled ‘flow control’ and the line from the left without label above the humidifier? If the butterfly symbolizes a MFC, why is the additional ‘flow control’ needed? G-CPC should be explained in the caption. Flow rate ranges (and pressure ranges) should be given for all flows, not only for one. Also the pump symbol should be in the legend for sake of completeness. A better match to ‘Humidifier’ would be ‘Dryer’ instead of ‘Dehydration’. The caption has an unmatched parenthesis.

110-125: I can’t make much sense of this section. It is too short to give a real explanation of what was done in this previous work of Bundke et al. 2015. And it doesn’t make sense showing a curve from the previous work and then stating, that another curve has ben used. Or is there a different physical meaning between xi and eta?

This section should be thoroughly reworked. Either extend it to give a short explanation what was actually done in which step, or remove it, refer to the literature (and in case, state what was different to the previous approach).

148-163: This seems to be one of the key achievements. But we learn here for the first time, that MAGIC is actually not specified for < 300 hPa, which might be on of the motivations of that study. Therefore, the problem should mentioned in the introduction. Also, some more information on the optimization procedure would therefore be useful (plots). E.g., at which laser voltages and which sensor thresholds / offsets the system operated with what efficiency?

From 155-157 we learn that detector offset and threshold are different properties, but the expressions are used before. Maybe a sketch of the instrument and its logic would help here following the explanation. Are offset and threshold applied to the same reference potential or do they apply to different part of the electronics?

165: Without further explanation, Fig. 5 should be in the method section describing the aerosol generation. What means ‘the particle mobility sizes were measured to 138 nm’? There is a size distribution displayed – the maximum of the soot distribution? Why is the size resolution as a result suitable for the cut-off characterization? Y axis: In Fig. 3, the symbol was used, but here a description instead of N_FCE. Unify (Applies also to other plots).

179: Fig. 6: The D50 apparently doesn’t match the fit curve. Which data does D50 refer to? Is the fit curve in this case then suitable? Same applies to Fig. 8

=====================================

Minor remarks/Corrections

39: Reference format

42: Reference format

44: ‘limited’: Chose another wording. Being a greenhouse gas might be unfavorable, but it not really a limitation (compared to a flammable material onboard aircraft). And water vapor is a greenhouse gas, too, though of course weaker.

46: Doesn’t the reference belong to the second statement after it?

70: Reference format

75: Remove ‘as well’?

101: What happened to 2.1 and 2.2?

102: Multiply charged?

103: It is not an error of the DMA, as it simply selects according to charge-to-size ratios (or effective mobility). It’s an error of the data interpretation by assuming a unique effective-mobility/size relationship. The effect of course is correctly described, but I suggest a more careful wording.

104: ‘This effect…’ These different sizes?

109: Figure 2 is not referenced (or referenced as Figure 3).  Figure 2 doesn’t add much information over the text. Remove.

112: While for N_FCE is quite clear, what it should be, the symbol is not explained above. This equation has no number, but the next one ha. Why?

121: ‘_’ before Xi

122: Reference format

130: ‘pressure detection’: barometer / pressure sensor is measuring

131: ‘until only the detector threshold is the only limit of signal detection’: unclear. Reword.

And why is the threshold decreased, if the laser power increases? One would expect that also the ‘background’ light intensity would increase, and therefore the threshold should be increased.

143: ‘optimised’ should be explained in the caption

144: ‘droplets, which need to be counted.’ ?

146: If there is a 1-sec average, what is the actually reading frequency?

146: 1-sec averaged à 1 s average or one-second average

146: It seems that [standard] laboratory  ?

171-173: Quirky. Rephrase.

173: corrected -> multiple-charge-corrected

174: concerning -> with respect to

174:remove  multi-charged

175: FEC -> FCE

179: different

181: The material ammonium sulfate should be mentioned in the text before the curves are discussed.

198: dryer

200: Fig. 8: The pressure levels should be sorted ascendingly. D50 is one time after, one time before fits.

210: Fig. 9: The pressure levels should be sorted ascendingly.

218: ‘square of the Pearson correlation coefficient’ or ‘coefficient of determination’ – but where is it?

234: affinity -> disinclination / repugnance

240: Table 2: Bb -> B ?

250-252: As long as there is not bypass sampling in used.

253: Bundke et al. 2015

257: Header has no number

260: ‘We recommend, testing’ remove comma

263: “It is noted that since this study, the manufacturer has modified…” That should maybe be noted with a remark at the according plots, otherwise a reader might overlook that the plots are no longer applicable to the current instrument generation.

265: So are the manufacturer setting acceptable now for this pressure range?

267: “Its well-engineered water recycling mechanism…” That information is new. Its relevance for the section is unclear.

266: “operates without loss in performance” The manuscript dealt with the details of exactly this performance loss, so this general statement doesn’t seem to be suitable for the conclusion section.

269: “To evaluate …” from here a summary start, which should be at the beginning of the last section.

276: Solubility is probably not a directly relevant property here.

277-278: “… Its well-engineered water recycling mechanism.” Unclear. Rephrase

278: “For pressures below 200 hPa, the efficiency of the MAGIC 210-LP can reach 100% linearity…” No data for this pressure range were shown in the manuscript, so the conclusion is a bit surprising. Or is this referring to pressure altitudes? Avoid ‘can’.

296: Some of the references don’t seem to be managed by a citation system, what would berecommended. Some dois are given as http-reference, some only numerically. Unify.

300: Details of publication missing?

303: doi is missing a ‘w’

Review of “Characterization of a self-sustained, water-based condensation particle counter for aircraft cruising pressure level operation” by Weber et al.

The paper details the performance of a COTS water CPC for use on the IAGOS aircraft. Instrument performance was compared to another butanol-based CPC and an electrometer throughout the pressure range of anticipated flight conditions for two different aerosol species.

The manuscript requires some clarification, added details, corrections, and further editing for grammar and punctuation. However, the discussion is suitable for the publication in AMT.

######################################################################################

Remarks/Questions

Abstract:

You say "simulated aircraft operational environment", but no temperature characterization across ambient range.  Is instrument sensitive to ambient temperature changes affecting sample temperature, and thus, supersaturation and cut size?

Saying "excellent agreement" between the instruments is misleading when you have performance differences in pressure for soot particles.

Figure 1:

Poorly drawn diagram.  Uneven spacing, crooked lines, random box sizes, critical orifice gap and protrusion.

Low pressure section incorrectly defined at Flow Control filter. Flow control and "dry side" of the humidity section are redundant.

With 4 flow controllers, do you have any measure of stability of the system?  How steady was the sample flow, pressure, and humidity control?

92-95: 

Why constant 30 second steps?  What were your statistics?  Your particle size distribution concentration is varying >3 orders of magnitude across the size range (fig 5), why hold constant DMA steps.  Increase time at small sizes to reduce massive error bars in counting statistics.

Why is no data shown above 60 nm if upper limit was 140 nm?

Is 2.5 nm lower limit corrected for diffusion losses changing distribution shape asymmetrically, and shifting peak upwards?  Any line loss analysis to approximate what the actual peak was when DMA is set to 2.5 nm?

What is your mixing chamber volume and flowrate to show that 15-seconds between samples is enough flush time?  Show flush time is at least 3-5*Tau.

101-125: 

Section labeling?  Whole section needs to be explained more thoroughly and clearly.

How is ksi determined? Is it calculated, determined experimentally?  Where is the equation for it?  How can one reproduce your correction method with the information provided here?

Figure 2: What flowrate and offset corrections?  Why and how are they performed? You have them listed in Figure 2 but never address.  Figure 2 does not add value.

148-163:

CPC operating parameters should be mentioned earlier (first introduction of the instrument).

This is your first use of the term "offset" without defining what it is or how it differentiates from the detector threshold.

You stated only two parameters are adjusted, laser power and detector threshold.  Now you're adjusting the offset too?

Move definition of offset at its first usage, and first introduce it when you're defining what parameters you adjust.

Move last sentence (162-163) to where you're talking about threshold set points.  You jump from threshold set, to laser power adjust, back to threshold set.

165:

State the aerosol types in the text.  Also why you used them as your reference aerosols.

Figure 5 and associated text should be in the Methods section.

Since your data is most significant in the 3-10 nm cut size range, use log y-scale so the concentrations used during the tests are more apparent.

Figure 6: No horizontal error bars accounting for DMA transfer function width.  What were your DMA flows?  Sizing accuracy analysis?

222-225:

If log-normal fit is inappropriate, don't use a log-normal fit.  Use the measured size distribution in your calculations.

Error between log-normal fit and measured size distribution affecting your multiple-charge correction can be calculated.  If you're using this as your explanation, prove it.

281-284: This is your first time discussing uncertainty. This should be discussed in detail in the results section, then summarized in conclusions.

######################################################################################

Minor remarks/corrections

11: Remove "and more"

26: Punctuation

31-33: nm particles can be detected via charging and electrometer, as you've used.  Suggest changing to "... growing them to optically-detectable droplets..."

38: Replace "by" a photodiode with "with" or "using"

51-53: Wordy. "However, butanol's flammability property strongly hinders..." Also, flammability does not hinder the operation, it hinders the desire to operate it.

54-59: Unbalanced parentheses and wordy.

60: define "low-pressure" range.  Can it be used in a balloon? High-altitude aircraft?

63-64: Incomplete sentence by itself. No subject.

64-65: define "broad pressure range" and define "aerosol types" and why.

71-73: Avoid using "it". Define. Rework sentence.

77:  Why is RH controlled to 30%.  Explain significance.  State where RH and temperature are measured.

102: Change to "multiply-charged particles".  What do you mean by measurement?  DMAs do not measure particle size, they size-select based on particle mobility.  The issue is using a mobility-based selector as an equivalent to a size-selector.

103: Replace "these" to avoid being ambiguous.  Hyphen singly-charged.

104: "this effect" is ambiguous.

105-106: "this artifact" ambiguous. To address multiply-charged particles biasing the concentration discrepancy...

111: Why is Multiple capitalized?

113: Why is Electrometer capitalized?

119-125: Mixture of fonts, inconsistent throughout text.  Assume document is printed B&W and can't refer to "red line".  Describe what the first order approx. means.

123: efficiency of what?

130-132: Confusing.

132-134: Merge sentences.

134: Compared to what temperature values at normal operation?

144-147: This section should be merged within the paragraph above

149:  Remove: initial.  Replace "to" with "at".

150: Missing comma, missing "is"

151: insert "and is shown..."

152: efficiently

153: comma after pressure.  "as a function"

154: lowered

157-158: merge sentences

158-159: Be specific and mention for the 250 hPa case...

159-160: redundant

174: suggest replacing "concerning" to "with reference to"

175: Comma after 7.

183: Not necessary to have "as illustrated in Figure 7". You've already stated you're referring to Fig 7. Sentences seem redundant with message.

194-195: This sentence seems out of place here... remove?

197: Explain why the second test aerosol case is necessary.  What are you exploring with the second choice of aerosol?

214: State this earlier in motivating your methods on why you chose this second case.

215: Source? There are many flights and missions targeting fresh combustion.  On commercial flight, you're flying in a corridor route that follows other aircraft. 

217: Refer back to Eqn 1 to remind the reader.   What is derived vs Exp?

218: As can be seen where?

220-221:  Are you saying that your method is incorrect?

222: remove "full-size"

234: reference.

Table 1 & 2:  Why is Table 2 Bb when Table 1 is B? Exp. stands for experimental or exponential?  Unclear.

243: Refer back to Eqn 1.

249: "agreement... is high". Be quantitative.  e.g. "agreement within 10% throughout the range..." "R^2 of ..."

253: Formatting

260: remove comma

261: insert comma after hPa. Change ot "as necessary"

262: "We were able to have a look at 5 units."  Relevance?

267: remove "well-engineered".  You're not in marketing.

270: change "was" to "were", since you're comparing 2 objects.

272: "Approved" is a strange word choice. Revise?

279: Should "below" be "above"?  You didn't test below 200 hPa.

A commercially available water-based CPC (MAGIC CPC ) is tested over a range of pressures to characterize its performance for use on aircraft, specifically IAGOS. The CPC is tested with both hygroscopic and hydrophilic aerosols and compared with a butanol CPC (GRIMM-CPC). The study, once the concerns mentioned below are addressed, represents a substantial contribution to scientific progress in enabling airborne measurement of aerosol concentrations without the use of toxic or high GHG potential working fluids.

The manuscripts describe in detail testing the water-based CPC in comparison to a reference electrometer and a Grimm-CPC over a pressure range 200-700 hPa. While much of the method and experimental set up are well described and justified, there are a number of aspects which need more full explanation or justification:

• Humified air appears to be added after the DMA, changing the humidity from 5 % to 30 %. Do the particles grow after the DMA size selection because of this? What was the residence time between the DMA and the CPCs and electrometer?
• In figure 1, what is flowing in through the two valves in the bottom section, and the flow controller in the top middle? Is it zero-air? Filtered lab air? Something else?
• What are the flow rates and line lengths between the DMA outlet and CPC and electrometer inlets? Were diffusion losses the same in all 3 lines?
• The authors discuss how the water-based CPCs respond differently to particles depending on their chemical composition, hence testing both soot and ammonium sulphate. It has previously shown that water-based CPC cut-off diameters can be substantially larger for organic aerosols e.g. (Hering et al., 2005) (by one of the co-authors). Why did the authors choose not to examine organics given their atmospheric relevance?
• Section 3 describes how the CPC increases laser power and decreases the detector threshold “until only the detector threshold is the only limit of signal detection” (page 5, line 132) . I think there may be a slight wording issue here, and beyond that, how is it ascertained when this limit is reached?
• Section 3, page 5, line 139 suggests that CPC overheating in hot ambient conditions could be addressed be maintaining a constant delta T between condenser and saturator, instead of trying to keep each at fixed settings. However, in the previous lines the need to keep the initiator below 45C and the condenser between 2 and 4C because of manufacturer specifications. Do these limits not mean that a constant delta T can therefore not be maintained?
• Section 3, page 5, line 146 mentions a look-up table for laser power as a function of ambient pressure. Was this something provided by the manufacturer or produced in this study?
• Section 3, page 6, line 155 to 164 discussed the need to vary the detector offset setpoint from 250 mV to 400 mV. It is not clear to me whether a single detector offset setpoint works for the full pressure range from ground to 200 hPa. Some clarification needed here.
• Fig 4 – why are there no error bars? The choice of 500 mW laser power seems well justified for the 250-700 hPa pressure range. Why was the lower limit of 200 hPa not tested here? Similarly, what about pressure greater than 700 hPa? Will this setting work to ground level? The authors mention that for 250 hPa and adjustment of the dectector offset was needed from 205 mV to 400 mV. Can this be done automatically in flight? Is there a loss of data while this change is made?
• Fig 5 shows an extended soot size distribution including measurements from another paper by the same 1^st Was this size distribution measured with the same experimental set up? If not, what makes it certain that the same size distribution occurred in these experiments as the other Weber et al. 2022 paper?
• Page 7 Line 181 claims “Using ammonium sulphate as a particle material, the instruments respond with an excellent agreement with the FCE reference instrument, with a slope of 1.0 ±0.05 regardless of the inline pressure.” Do the authors mean that a linear fit was performed for each pressure, and all of them had a slope of 1.0 ±0.05 individually? Or was the fit made of the aggregate data over all pressures? It needs to be clearer what was done. If only the aggregate fit was done, the data at higher pressures might make the agreement seem more robust than it really is at low pressure. This graph also lacks uncertainties. Later, around line 220 the authors discuss that both CPCs see up to 15% fewer particles than the electrometer. This seems to contradict the “excellent agreement with the FCE reference instrument”.
• Section 3, page 9, line 220 discusses that the undercounting of the CPCs relative to the electrometer for soot particles may be due to an error in the multiple charge correction because the size distribution was not measured above about 150 nm. But in fig 5, the size distribution is presented using a different study up to 1mm and clearly covering the bulk of the generated mode. Can this extended size distribution not be used to better calculate the multiple charge correction and correct this undercounting?
• In the conclusions, page 11 line 262, it is mentioned that 5 units were examined. This is the first mention of this (would be better in the methods description). It is not clear whether the data presented are from all 5 units, or just 1 CPC. This needs clarification, and the claim that “factory settings for all 5 units” were satisfactory down to 200 mb needs to be supported by data.
• Section 3 analyses a number of different aspects of instrument performance, and clarity might be improved by introducing subsections.

Many of the conclusions drawn in this manuscript are well justified, but the following do not seem well supported:

• The experiments performed in this study cover the pressure range of 700-200 hPa. Why was 700 hPa chosen as the lower limit? The title suggests that the study only is interested in “cruising pressure levels” in which case the lower cut off of 700 hPa is well justified. This should be made clearer also in the text. The question of instrument performance between 700 hPa and ground seems to be left unanswered.  The conclusion, page 11 line 278 claims “This pressure range covers the operational conditions present during IAGOS aircraft flights.” Do IAGOS flights only use data at pressures lower than 700 hPa?
• Section 3 line 205 claims fig 8 shows the counting efficiencies of the MAGIC CPC and GRIMM CPC are nearly identical for pressure above 250 hPa. This is hard to see from the graphs, and from the fits shown, they seem to have different cut off diameters, so then the behaviour is quite different. I’m not sure what was meant here, and it may be that different plots are needed to show it well.
• Figure 8 – it is very hard to distinguish the multiple shades of green or blue from each other in these plots, both for the points and the error bars. It makes it hard to determine if the conclusions drawn e.g. about the stability of the G-CPC efficiency curve over the examined pressure range, are valid. The order of the legends makes the plots very confusing to read. The D50 lines are very confusing. They appear to not pass through any of the presented fits. How are they determined? And why do they not pass through the fit lines?
• Fig 9 appears to show a trend with ambient pressure of the CPC number concentrations to that measured by the electrometer, but this is obscured by the choice of a linear scale, the hard-to-distinguish colours, the lack of error bars and the confusing ordering of the legend. Until this is corrected, it is not possible to determine whether the claims made around line 205 are justified.
• Page 11 line 249 “Overall, the agreement between values derived directly from the experiment and values deduced from the fitting procedure is high.” What objective metric is this based on?
• Conclusions, page 11 line 265 “the manufacturer has modified the firmware and design of the MAGIC 210-LP to improve the performance at high altitudes and to better accommodate the automatic adjustments in the laser and detector settings with operating pressure.” It is unclear what aspects of “performance” have been enhanced since this study was made, and also which of the laser power and offsets that were tested and adjusted in this study are now different in the instrument settings. The whole concept of what works with the CPC settings and what needs adjusting by the user, and what is adjusted automatically vs needing to be adjusted by the user as the pressure changes needs to be much more clearly addressed.
• Conclusions, page 11 line 258 “excellent overall performance compared to a standard butanol CPC” need to be more specific. The shift in D50 with hygroscopicity should be mentioned, also the shifting D90 with changing pressure.

Some of the terminology, wording and presentation used is confusing to the reader, and needs adjusting:

• Abstract, line 21 “the D50 cut-off diameter did not differ significantly for particle sizes around 10 nm” – does this mean the D50 cut-off diameter was stable at around 10 nm over the tested pressure range?
• Section 3, line 182 references the “Sky CPC”, which is otherwise referred to as the G-CPC. It would be clearer to use consistent terminology.
• Introduction, line 26 “adverse effects that particles can have on climate change” seems a bit confused, suggest revision. It makes more sense with reference to air quality, which is listed later.
• Introduction, line 54 “It comprises”, unclear if referring to this study or IAGOS. Suggest it is better to be specific than let the reader infer from the following text.
• Section 3 line 154: “lowed”?
• Section 3, line 207 describes the increase in CPC cut off for soot particles as occurring “as soon as an ambient pressure of 200 hPa is reached”, this makes it sound like a sudden transition, whereas in reality it will be a gradual shift with decreasing pressure. The text should be changed to more accurately reflect this.
• Page 9 line 226, the sentence “Looking at the D50 Value of Tables 1 and 2, both Instruments have the same range of particle diameter of 5 nm for ammonium sulphate” is unclear and needs revision
• Tables 1 and 2 – what are A, B and Bb? These do not seem to be defined anywhere.
• Conclusions, page 11 line 259 “LP” acronym needs definition, and why is it only introduced here and not earlier?

Overall relevant work seems well referenced. I suggest the following additions/corrections:

• Introduction, line 42 – the more relevant reference is again Williamson et al. (2018), which describes the referenced flourinert based CPC in detail, as opposed to a combined payload including that CPC that is described in the current reference.

References:

Hering, S. V., Stolzenburg, M. R., Quant, F. R., Oberreit, D. R., & Keady, P. B. (2005). A Laminar-Flow, Water-Based Condensation Particle Counter (WCPC). Aerosol Science and Technology, 39(7), 659-672. https://doi.org/10.1080/02786820500182123

Williamson, C., Kupc, A., Wilson, J., Gesler, D. W., Reeves, J. M., Erdesz, F., McLaughlin, R., & Brock, C. A. (2018). Fast time response measurements of particle size distributions in the 3–60 nm size range with the nucleation mode aerosol size spectrometer. Atmos. Meas. Tech., 11(6), 3491-3509. https://doi.org/10.5194/amt-11-3491-2018