Characterizing the performance of a POPS miniaturized optical particle counter when operated on a quadcopter drone

Liu, Zixia; Osborne, Martin; Anderson, Karen; Shutler, Jamie D.; Wilson, Andy; Langridge, Justin; Yim, Steve H. L.; Coe, Hugh; Babu, Suresh; Satheesh, Sreedharan K.; Zuidema, Paquita; Huang, Tao; Cheng, Jack C. H.; Haywood, James

doi:https://doi.org/10.5194/amt-14-6101-2021

Articles | Volume 14, issue 9

https://doi.org/10.5194/amt-14-6101-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/amt-14-6101-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 14, issue 9

Research article

|

16 Sep 2021

Research article |

| 16 Sep 2021

Characterizing the performance of a POPS miniaturized optical particle counter when operated on a quadcopter drone

Zixia Liu, Martin Osborne, Karen Anderson, Jamie D. Shutler, Andy Wilson, Justin Langridge, Steve H. L. Yim, Hugh Coe, Suresh Babu, Sreedharan K. Satheesh, Paquita Zuidema, Tao Huang, Jack C. H. Cheng, and James Haywood

Download

Final revised paper (published on 16 Sep 2021)
Preprint (discussion started on 21 Jan 2021)

Interactive discussion

Status: closed

RC1:
'Comment on amt-2020-495', Anonymous Referee #1, 22 Jan 2021

Overall Comments:

The authors present a ground-based comparison of a POPS measurement and SMPS in an environment heavily influenced by biomass burning aerosol. The authors then compare POPS measurements on a UAS in flight to its measurements on the ground and suggest that these differences are due primarily to the UAS flight.

Although comparing an SMPS to the POPS is not particularly novel, this study does verify that the agreement holds in an environment influenced by biomass burning and where a different index of refraction has been assumed.

The authors suggest that the POPS is adversely affected in this particularly sampling position on a small rotary UAS. Yet the authors note that no attempts were made to sample from other locations on the multi-rotor aircraft, or to modify the inlet for in-flight aerosol measurements. Potential perturbation to instrument flow are not directly addressed. These should perhaps be persued. At the very least, the authors must tailor the scope of their claim to say that simplistic measurements form the POPS instrument are ill advised and quad copter aerosol sampling requires careful evaluation and consideration.

The flight tests do not involve an in-flight inter-comparison, which we suggest the authors also pursue, if possible. As noted, real differences in aerosol distributions and particle concentration number with time could have obscured perceived UAS aerosol sampling bias.

Specific Comments:

The abstract should be shortened.

L31 – L33 An odd comment; remove from the abstract. Begin with, “we compared the Portable Optical Particle Spectrometer, a small light-weight and high sensitivity optical particle counter…”

L37 Awkward. Rephrase. “This is the first such documented…

L38 Word choice – you don’t “investigate” the RMSE or MAD - you report it.

L50 – 52 Be specific about differences in coarse mode to what other instrument – SMPS does not measure particles > 1.0 micron. This is particularly confusing…

L62-74 This section could be much more concise.

L88 Awkward phrasing.

L106 – 113 This section could be shortened and only details particularly relevant to this study should be mentioned (this overall description is covered in Gao et al. 2013 and 2016).

L117 Did this study include a POPS?

L119 Comparisons to tower measurements – what instruments were compared and were they compared only at one height? Was temporal averaging applied?

L131-132 It is still not clear what was entailed in the in-flight UAS POPS comparison. Please be more specific.

L132-134. This sentence should be removed. It is not helpful.

L153-155 Was the adjustment to account for a difference in the index of refraction done to binned data or per particle data? Doing this to binned data could introduce an additional (likely small) source of error.

L178-180 Can the authors comment on how the sampling tube might be optimized for drone sampling? This seems like a very important point considering the comparison/ test.

L205 Was this date of the wing-mounted PCASP instrument a day that the POPS sampled (on the UAS or on the ground)? What altitudes were sampled to provide these size distributions?

If not, perhaps shorten this section and specify that the PCASP size distribution is simply provided for reference.

L235-240 For an instrument comparison, the size range of the SMPS and POPS should only be compared in the range where measurements overlap. The full SMPS and POPS size ranges should only be used to characterize atmospheric aerosol distributions more fully.

L258 – 290. Since no in-flight comparison to another instrument was done, the authors need to demonstrate that they did not observe any systematic differences in the PSD at 10 m compared with at the ground.

L285-287 not needed.

L301-303. This is a good point. The POPS flow (used to calculate PC in each bin size) needs to be monitored in each of the different flight positions.

L313 – 316. Again what is the flow in each POPS case? Are the counting statics much poorer during FLY than when on the ground?

L331-333 Unfortunately, this point undermines this entire study. If there are real differences that might be confused with instrument performance in flight, the study ideally should address this.

L336-337 This does not make sense. Two additional stages of G_NR are suggested?

L356- end Tailor sweeping claims as suggested.

Citation: https://doi.org/10.5194/amt-2020-495-RC1
- AC1: 'Reply on RC1', Zixia Liu, 08 May 2021
  
  We would like to thank the reviewers for their comments on the manuscript. The reviewers raised some interesting and relevant points that we have addressed in our revised manuscript. Our responses are presented in the document (see supplement) in red.
  
  Citation: https://doi.org/10.5194/amt-2020-495-AC1
RC2:
'Comment on amt-2020-495', Anonymous Referee #2, 01 Mar 2021
This manuscript investigated the performance of POPs on a drone and compared with aerosol microphysics measurements observed at the surface. This study provides a series of methods to quantify the uncertainties of the PNCs and PSD measured onboard a drone. The results from this study provide a good reference to the community when one wants to use the POPS or similar instruments to access aerosol profiling issues. The manuscript is well written in English and content structure. My comments are listed bellows:

Higher wind speed increased measurement error. The author explained that it is caused by the drone suffers from variations in pitch and yaw. It is not clear to the reviewer. If the wind direction is fixed during the flight, how the drone changes pitch and yaw frequently. My personal opinion is that the inlet flow rate of POPS is small, the high speed of horizontal wind causes the measurement uncertainty due to the insufficient inlet flow.

In the abstract (line 59), the measurement errors induced from turbulence need to be carefully stated since the authors did not provide turbulence measurement data to support the conclusion.

Introduction: The first paragraph is too long and can be divided into three paragraphs.

Line 138: please check latlon format.

Lines 157-160: not clear to reviewer, rewritten the sentence is needed.

Line 278: T14 was marked as red in the table.

Lines 250 and 351. The values from what kind of surface measurements should be specified.

In Figure 3. The unusual spike values (bad data) may remove from the figure. X-axis label should be improved.

In Figure 5: The dark blue line can be changed as green.

Table 5: All flights à all cases. The PNC MAD (%) might merge to the same raw of PNC RMSD(%).

Table 6 might present in two tables, same as tables 4&5.
Citation: https://doi.org/10.5194/amt-2020-495-RC2
- AC2: 'Reply on RC2', Zixia Liu, 08 May 2021
  
  We would like to thank the reviewers for their comments on the manuscript. The reviewers raised some interesting and relevant points that we have addressed in our revised manuscript. Our responses are presented in the document (see supplement) in red.
  
  Citation: https://doi.org/10.5194/amt-2020-495-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Zixia Liu on behalf of the Authors (25 May 2021) Author's response Author's tracked changes Manuscript

ED: Publish subject to minor revisions (review by editor) (15 Jun 2021) by Troy Thornberry

I would like to thank the authors for their efforts in addressing the comments on their originally submitted manuscript that were made by the peer reviewers. I think the manuscript presents a reasonable analysis of the data the authors have collected on the performance of their POPS instrument in both a ground comparison with a SMPS reference instrument during the CLARIFY 2017 campaign and by comparing flight (hover) to non-flight (ground) performance when mounted on a quadcopter UAS.

I would encourage the authors to address a few additional comments prior to publication:

L40: perhaps “operating on the ground and when hovering at 10 m.” since both are with the POPS mounted on the UAS.

L60: “acting as condensation nuclei” -> “to act as cloud condensation nuclei”

L109: “use the” -> “used a” or “included a”

L138: SMPS sample was dried while POPS was not. How close was the POPS temperature to the ambient sampling temperature?

L150: it would be useful for readers if the authors would add some detail to the process used to adjust the sizing (signal -> diameter) for the change in refractive index from PSL to that assumed for biomass burning. Was the Mie calculation (polarized?) carried out for the POPS light collection geometry or more generally? Mie code used for calculation?

L161: SMPS range is stated as 0.01 µm to 1.00 µm, but in Fig 2 and the description of the overlap size range the upper limit is 0.45 µm.

L180 (Table 1): Table 1 as constructed seems unnecessary. The dates are already included in Table 3 along with other important information—perhaps Table 1 could be replaced by Table 3 with an additional column for time.

L209: “contribute to 93% of the scattering” should be just “contribute 93% of the scattering”

L211-: it would be useful to include in the text some of the details that appeared in your response to reviewer 1’s question about the CLARIFY PCASP data that are shown in Fig 2: POPS and SMPS were sampling at 330 m while the BAE146 PCASP measurements were 1.9-7.3 km and the potential causes of the divergence at larger Dp.

L239: “mass mixing ratio”—Figure indicate volume mixing ratio (ppmv)

L251->: it seems that the typical (although not universal) increase in PNC between G_R and G_NR could include a contribution from suspension of particulate material by the near-ground prop wash (also coarse more impacted than accumulation?), but the same is not true for the even larger typical increase between FLY and G_NR. What is the proposed mechanism by which higher wind speeds lead to artificially higher PNCs in flight?

L293: “Variations in the orientation of the inlet led to uncertainties in the sample flow rate.” Is vague. The POPS flow and uncertainty in the flow are independent of instrument orientation. One might postulate based on the observed increase in the scatter of the reported flow (whether real or perceived) during flight (FLY relative to G_R), that the fast adjustment of the platform to changes in wind produces this variability in the flow. The noise in the flow does not increase appreciably from G_NR to G_R, but there is an observed mean increase in PNC.

L302: How (though what mechanism) do you expect the noise in the reported flow impacted the PNC measurement over time? L299 states that the mean flow rate was unchanged, but the result seems to be a systematic increase in the measured PNC.

L309: “differences at sub-micron sizes are less than those at” or “differences in sub-micron aerosol are smaller than those in super-micron aerosol”

L312: “summarizes the PSD percentage differences for the two modes”

L327: What is meant by POPS “operated in” accumulation mode vs coarse mode? Are these not just different parts of the size distribution (per line 311)?

L391: What would the hypothetical mechanism for the rotor induced increase in coarse mode aerosol be?

Figure 3 caption: perhaps “(a) Time series of SMPS and POPS particle concentrations in the diameter range 120 – 450 nm measured during the LASIC/CLARIFY-2017 campaign. (b) Ratio of the POPS to SMPS concentrations shown in (a).” I agree with the spikes in the CO trace from the daily calibrations is distracting. In panel (e), it would be better to not include the lines between gaps in the AOD data.

Figure 4: it would seem that using a consistent binning across the G_NR, G_R and FLY PDFs of each flight would provide a better comparison of the respective PNC distributions than the variable bin widths that are presented.

Hide

AR by Zixia Liu on behalf of the Authors (08 Jul 2021) Author's response Author's tracked changes Manuscript

ED: Publish as is (24 Jul 2021) by Troy Thornberry

AR by Zixia Liu on behalf of the Authors (02 Aug 2021)

Short summary

This paper first validates the performance of an advanced aerosol observation instrument POPS against a reference instrument and examines any biases introduced by operating it on a quadcopter drone. The results show the POPS performs relatively well on the ground. The impact of the UAV rotors on the POPS is small at low wind speeds, but when operating under higher wind speeds, larger discrepancies occur. It appears that the POPS measures sub-micron aerosol particles more accurately on the UAV.