A model for simultaneous evaluation of NO2, O3, and PM10 pollution in urban and rural areas: handling incomplete data sets with multivariate curve resolution analysis

Gorrochategui, Eva; Hernandez, Isabel; Tauler, Romà

doi:https://doi.org/10.5194/acp-22-9111-2022

Articles | Volume 22, issue 13

https://doi.org/10.5194/acp-22-9111-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/acp-22-9111-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 22, issue 13

Research article

|

14 Jul 2022

Research article |

| 14 Jul 2022

A model for simultaneous evaluation of NO₂, O₃, and PM₁₀ pollution in urban and rural areas: handling incomplete data sets with multivariate curve resolution analysis

Eva Gorrochategui, Isabel Hernandez, and Romà Tauler

Download

Final revised paper (published on 14 Jul 2022)
Supplement to the final revised paper
Preprint (discussion started on 20 Apr 2022)
Supplement to the preprint

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2022-117', Anonymous Referee #1, 06 May 2022

This study attempts to apply multivariate curve resolution alternating least squares with quadrilinearity constraints to handle complex and incomplete four-way atmospheric data sets. Three air pollutants including nitrogen dioxide, ozone and particular matter in eight sampling stations located in Barcelona metropolitan area other parts of Catalonia during the COVID-19 lockdown (2020) with respect to previous years (2018 and 2019) are used for analysis. The new method indeed generates an interesting result, which has been well interpreted. Overall, the manuscript is well written and worthy of publication. However, this reviewer disagrees on one issue, i.e., estimation of missing data always suffered from uncertainties, whatever any approach to be used. The uncertainty should be included in data interpretation.

Citation: https://doi.org/10.5194/egusphere-2022-117-RC1
- AC1: 'Reply on RC1', Eva Gorrochategui, 10 May 2022
  
  We thank the reviewer for the comment. And yes, we agree that measurement uncertainties were not included in the bilinear model factor decomposition estimations. The environmental agency source data did not provide them. Otherwise, we could have applied the weighted version of ALS where data uncertainties are included as weights in the least squares estimations. On the other hand, missing data blocks were not included in the least squares estimations, this is the advantage of the proposed method, linear equations were only solved for the known data blocks. Therefore this should not be a limitation. What is true is that some parts of the factor solutions (those corresponding to the missing blocks) are not so overdetermined from a least squares point of view as the other data blocks without missing values, and this can be reflected in the reliability of the estimations of the later. This is an aspect that deserves a deeper study and needs further investigation.
  
  Citation: https://doi.org/10.5194/egusphere-2022-117-AC1
RC2:
'Comment on egusphere-2022-117', Vasil Simeonov, 12 Jun 2022

The present study proves convincingly the advatage of the chemometric technique multivariate curve resolution - alternating lesdt squares. The reasons for such an estimations could be summarized as follows:

1. Reliable options for apportionment of pollution sources

2. Realing with four-way constructed data sets

3. Elimination of missing data

The study is performed on a very high theoretical level and, additionally, delivers very useful practical information. The style is very sound, the conclusions - important.

I recommend acceptance in its present form.

Citation: https://doi.org/10.5194/egusphere-2022-117-RC2
- AC2: 'Reply on RC2', Eva Gorrochategui, 13 Jun 2022
  
  We kindly appreciate this reviewer's comment. We also agree that this study highlights the advantages of chemometrics in the analysis of complex four-way environmental data sets containing missing data blocks.
  
  Citation: https://doi.org/10.5194/egusphere-2022-117-AC2

Peer review completion

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

AR by Eva Gorrochategui on behalf of the Authors (13 Jun 2022) Author's response Manuscript

EF by Una Miškovic (15 Jun 2022) Author's tracked changes

ED: Publish as is (21 Jun 2022) by Leiming Zhang

AR by Eva Gorrochategui on behalf of the Authors (22 Jun 2022)

Download

Article (3993 KB)
Full-text XML

Short summary

A multiway methodology is proposed to handle complex and incomplete atmospheric data sets, providing concise and easily interpretable results. Changes in air quality by NO₂, O₃ and PM₁₀ in 8 sampling stations located in Catalonia during the COVID-19 lockdown with respect to previous years (2018 and 2019) are investigated. Simultaneous analysis of the 3 contaminants among the 8 stations and for the 3 years allows the evaluation of correlations among the pollutants, even when having missing data.