Performance evaluation of the Alphasense OPC-N3 and Plantower PMS5003 sensor in measuring dust events in the Salt Lake Valley, Utah

Kaur, Kamaljeet; Kelly, Kerry E.

doi:https://doi.org/10.5194/amt-16-2455-2023

Articles | Volume 16, issue 10

https://doi.org/10.5194/amt-16-2455-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Special issue:

Dust aerosol measurements, modeling and multidisciplinary...

https://doi.org/10.5194/amt-16-2455-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 16, issue 10

Research article

|

24 May 2023

Research article |

| 24 May 2023

Performance evaluation of the Alphasense OPC-N3 and Plantower PMS5003 sensor in measuring dust events in the Salt Lake Valley, Utah

Kamaljeet Kaur and Kerry E. Kelly

Download

Final revised paper (published on 24 May 2023)
Supplement to the final revised paper
Preprint (discussion started on 02 Dec 2022)
Supplement to the preprint

Interactive discussion

Status: closed

CC1:
'Comment on amt-2022-292', JAMES OUIMETTE, 05 Dec 2022

This evaluation of the OPC-N3 for PM10 during wind blown dust episodes is both timely and valuable. FYI, the Great Basin APCD headquartered in Bishop, California, is starting to evaluate the OPC-N3 sensor within the Alphasense monitor for wind blown dust PM10 at Keeler, next to the Owens dry lake, and at Mono Lake. I have contacts if you're interested.
A minor correction to your preprint - - on lines 61-63 it states that some low cost sensors are ineffective at measuring PM10 and dust, primarily due to the sensor’s inability to aspirate these larger particles into the device, and cites our paper by Ouimette (2022). In our paper we did not state that poor aspiration efficiency was the primary reason for the coarse particle inefficiency. Based on the Plantower PMS5003 particular geometry our physical-optical model predicted an 80-90% reduction in the 2-um particle scattering coefficient compared to a perfect nephelometer, due to truncation of the forward scattering signal. However, the lab data showed more like a 95% reduction in 2-um light scattering. Based on Willeke and others' work we hypothesized that poor aspiration efficiency could be a reason for this additional 5-15% loss. But we neither modeled nor measured it for that paper. It's extremely difficult to measure aspiration losses for the PMS5003 because any measurement slows or stops its flow rate due to Δp across the little fan.
I hope this helps. Best wishes on your paper review.
Regards,
Jim Ouimette

Citation: https://doi.org/10.5194/amt-2022-292-CC1
- CC2:
  'Reply on CC1', Kerry Kelly, 09 Dec 2022
  
  Thank you very much for the kind words and this correction. We will ensure that this text modified in the final version. We also appreciate the connection to the group working on Owens Lake dust and will drop you a separate note about that.
  
  Citation: https://doi.org/10.5194/amt-2022-292-CC2
  - RC1: 'Reply on CC2', Shahrul Nadzir, 23 Dec 2022
    
    Dear Kelly et al
    
    Thank you for your answer to the comment by the public. I have read through the entire manuscript and it has been well-revised.
    
    Citation: https://doi.org/10.5194/amt-2022-292-RC1
    
    AC2: 'Reply on RC1', Kerry Kelly, 19 Feb 2023
    
    We thank the reviewer for taking the time to consider our manuscript and for the kind words.
    
    Citation: https://doi.org/10.5194/amt-2022-292-AC2
RC2:
'Comment on amt-2022-292', Anonymous Referee #3, 03 Feb 2023
Overall, this is a well written paper that explores the performance of 2 common air sensors for measuring PM10. This topic is of great interest to many air sensor users as many inexperienced users are using these devices without knowledge of their limitations for PM10. I have a number of minor comments and suggestions.
Multilinear regression with RH: Is this the correct form of equation to look at the influence of RH? It seems like past work with the Alphasense OPC has found nonlinear influence of RH (e.g., https://amt.copernicus.org/articles/11/709/2018/). How did you decide to use 85% as your cut off?

Limitations Line 480: I think it would be important to mention that the lack of significant improvement of adding RH was after excluding the high RH data.

Lines 369-372: Did you consider if any of your PMS5003s were the “new” version of the PMS5003 and if that lead to any of the differences in agreement https://community.purpleair.com/t/new-version-of-plantower-pms5003/288/25, https://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=355518&Lab=CEMM&simplesearch=0&showcriteria=2&sortby=pubDate&searchall=barkjohn&timstype=&datebeginpublishedpresented=02/03/2021

Did you consider whether the particle size bins provided by the PMS5003 provided any better estimate of PM10? (e.g., https://www.mdpi.com/1424-8220/22/13/4741, https://www.mdpi.com/1424-8220/22/7/2755, https://amt.copernicus.org/preprints/amt-2022-265/)

Table 2: It might be helpful to add a row for the non-dust impacted times for easy comparison. Same suggestion for Table S1.

Consider adding averaging interval to all figure labels in text and SI (Assuming they are all 1-hr?)

Fig S1: Coloring points with >85% RH would be helpful

Introduction: Recommend adding the US EPA standard for PM 10

Typo in line 46
Citation: https://doi.org/10.5194/amt-2022-292-RC2
- AC1:
  'Reply on RC2', Kerry Kelly, 09 Feb 2023
  Multilinear regression with RH: Is this the correct form of equation to look at the influence of RH? It seems like past work with the Alphasense OPC has found nonlinear influence of RH (e.g., https://amt.copernicus.org/articles/11/709/2018/). How did you decide to use 85% as your cut-off?
  
  Response: We thank the reviewer for pointing this out. We did follow the AMT article suggested by the reviewer for deciding the relative humidity cut point and the multi-linear regression. According to Crilley et al. (https://amt.copernicus.org/articles/11/709/2018/), the non-linear behavior (the exponential behavior) of the OPC-N2 with relative humidity was observed for relative humidity greater than 85%. They indicated that for relative humidity less than 85%, a linear response could be observed between then OPC response and relative humidity. This linear behavior for relative humidities below 85% was also evident from the figures in Crilley et al. article. Therefore we performed a multi-linear regression (excluding measurements with relative humidity > 85%) with FEM-HW PM10 as the dependent variable and OPC-HW (OPC at the Hawthorne monitoring station) and relative humidity (RH) as independent variables.
  Limitations Line 480: I think it would be important to mention that the lack of significant improvement of adding RH was after excluding the high RH data.
  
  We thank the reviewer for pointing this out. We have corrected the statement to include the relative humidity cut point as a limitation. Corrected statement: “However, this study excluded measurements with a RH > 85% (<2% of total measurements), a range where previous studies have identified a significant effect of RH (Crilley et al., 2018), and the applicability of this study’s results to other, more humid, regions would need to be evaluated.”
  Lines 369-372: Did you consider if any of your PMS5003s were the “new” version of the PMS5003 and if that lead to any of the differences in agreement https://community.purpleair.com/t/new-version-of-plantower-pms5003/288/25, https://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=355518&Lab=CEMM&simplesearch=0&showcriteria=2&sortby=pubDate&searchall=barkjohn&timstype=&datebeginpublishedpresented=02/03/2021
  
  We thank the reviewer for pointing this out. We were aware of this issue, and we have confirmed with PurpleAir that all the PMS sensors used in this study were old versions.
  Did you consider whether the particle size bins provided by the PMS5003 provided any better estimate of PM10? (e.g., https://www.mdpi.com/1424-8220/22/13/4741, https://www.mdpi.com/1424-8220/22/7/2755, https://amt.copernicus.org/preprints/amt-2022-265/)
  
  We did not attempt to calculate the PM mass from the numbers as we didn’t know the density of the particles. Also, from our previous work and also reported by other researchers, the Plantower sensor does not respond to particles bigger than 2.5 µm.
  Table 2: It might be helpful to add a row for the non-dust impacted times for easy comparison. Same suggestion for Table S1.
  
  We have added the non-dust times in both tables.
  Consider adding averaging interval to all figure labels in text and SI (Assuming they are all 1-hr?)
  
  We have added the statement describing the averaging interval, i.e., 1 hour, in all of our figures including supplementary material.
  Fig S1: Coloring points with >85% RH would be helpful
  
  We have marked the high humidity point as red in Fig. S1.
  Introduction: Recommend adding the US EPA standard for PM 10
  
  We have added the statement about the EPA standard for PM10 in the introduction. The statement added “ The US EPA’s national ambient air quality standard for PM₁₀ concentration and are 150 and 50 µg/m³ for the 24-hour and annual average, respectively”
  Typo in line 46
  
  We have resolved the typo.
  
  Citation: https://doi.org/10.5194/amt-2022-292-AC1

Peer review completion

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

AR by Kerry Kelly on behalf of the Authors (02 Mar 2023) Author's response Author's tracked changes Manuscript

ED: Publish as is (25 Apr 2023) by Paolo Laj

AR by Kerry Kelly on behalf of the Authors (28 Apr 2023) Author's response Manuscript

Post-review adjustments

AA: Author's adjustment | EA: Editor approval

AA by Kerry Kelly on behalf of the Authors (22 May 2023) Author's adjustment Manuscript

EA: Adjustments approved (22 May 2023) by Paolo Laj

Short summary

We evaluated the AlphaSense OPC-N3 and PMS5003 compared to federal equivalent method (FEM) PM₁₀ measurements in the Salt Lake Valley during five dust events. Before correction, the OPC-N3 agreed well, but the PMS PM₁₀ measurements correlated poorly with the FEM. After correcting the PMS with a PM_2.5 / PM₁₀ ratio-based factor, the PMS PM₁₀ correlations improved significantly. This suggests the possibility of better resolved spatial estimates of PM₁₀ using PMS measurements and PM_2.5 / PM₁₀ ratios.