the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 14 May 2025
Assessing Spatial and Temporal Urban Air Quality Variabilities with the Vaisala AQT530 Monitor
Abstract. Low-cost gas and particle sensors can significantly increase the spatial coverage of Air Quality (AQ) monitoring networks in urban settings. Considering that the accuracy of such sensors is not high enough to replace reference instruments for AQ monitoring, the question is whether they can be used to capture spatial differences among different stations, as well as temporal trends and month-to-month variabilities at a specific location. To investigate that, we carried out measurements over a period of 19 months with two Vaisala AQ Transmitters-Monitors (Model AQT530), collocated with reference-grade instruments, in two AQ monitoring stations in Nicosia: an urban traffic and an urban background station. The AQ monitors employ Low-Cost Sensors (LCSs) for gaseous pollutants (i.e., CO, NO2, NO, and O3) and Particulate Matter (PM). Statistical analysis of the reference measurements shows that the mean concentrations of the pollutants at the two stations, determined over the entire study period and for each month separately, were significantly different. Analysis of the LCS measurements showed that that the reproducibility of the NO2, NO, O3, and PM2.5 sensors, over a period when these were co-located at the traffic station, is poor, excluding them from further investigating their ability to capture spatial differences between different stations. The CO and PM10 measurements from the AQ monitors effectively captured the differences in pollutant concentrations between the two stations when averaged over the entire study period or on a monthly basis, with few exceptions during specific months depending on the sensor. These LCSs were also able to capture concentration differences between the two stations on a daily or monthly basis, as long as those were above a certain threshold for each pollutant. The CO and PM sensors captured the month-to-month trend over the entire period of the measurements, similarly to the reference instruments, while the NO2, NO and O3 sensors did not, mainly due to their sensitivity to the environmental conditions. Despite that, all sensors captured the statistical significance of the month-to-month concentration differences at the same station, with the PM2.5 measurements showing the highest capability of doing so in accordance with the reference instruments.
Status: open (until 08 Jul 2025)
-
CC1: 'Comment on egusphere-2025-1749', Priyanka DeSouza, 03 Jun 2025
reply
Major comments
The authors compare the performance of two LCS and a reference monitor- specifically they evaluate temporal and spatial trends and find somewhat of a correlation. Why was no calibration model applied, despite the Introduction stating that such models are necessary? Why were the sensors not co-located with the reference monitor for a longer time to compare the two? Doing so would provide much more definitive results than the authors current approach. The authors results are not novel. I would advise the authors to use the sensors to investigate a specific air quality issue, instead of merely providing correlations between these sensors
In addition, a lot of key details are in the SI such as details of the WRS test.
Minor comments
In section 2.2 was there any reason the Vaisala monitors were selected? Have these sensors been evaluated by the EU JRC or CARB or US EPA? Can the authors report what these organizations have found during their lab calibrations/field tests?
In section 2.3 can the authors provide information on the manufacturer specifications
Citation: https://doi.org/10.5194/egusphere-2025-1749-CC1 -
RC1: 'Comment on egusphere-2025-1749', Anonymous Referee #1, 05 Jun 2025
reply
“Assessing Spatial and Temporal Urban Air Quality Variabilities with the Vaisala AQT530 Monitor”
General Comments
The manuscript presents the assessment of the Vaisala AQT530 Monitor in Cyprus. It is a long-term study (19 months) of two units deployed in an urban background and traffic location which has reference grade instrumentations for the co-monitored pollutants (CO, NO, NO2, O3, PM2.5 and PM10). The authors showed that only the CO and to lesser extent PM readings had reasonable agreement with the reference monitoring data in terms of temporal (diurnal, seasonal) and spatial (background vs traffic) variability. They attributed the poor performance of the Vaisala AQT530 for the other species to the impact of environmental factors like RH and temperature. The strength of this work is the detailed statistical analysis carried out by the authors during the assessment of the AQT530.
Specific comments
The reviewer if of the opinion that this work shows the performance limitation of the commercial LCS device that the authors used rather than a general conclusion of the limitations of LCS application in ambient AQ monitoring because there are large body of work that shows this, the authors need to capture this in some form in their conclusions.
Technical corrections
The abbreviation ‘cf.’ is incorrectly being used throughout the manuscript. It means “to compare to” and I think the authors are using it to mean “see”. Authors need to correct this.
Authors sometimes also make statements that are not really justified in the first instance but are subsequently justified later on in the manuscript. For instance, in the final paragraph on page 6, the authors introduced periods classified as dust and non-dust period without justification or any citation to back this up, but did subsequently add a citation on page 8 second sentence when discussing the dust event. I would suggest the authors look into instances of this and make corrections.
On page 8, the reviewer also struggle to agree with authors the general conclusion that systematic bias in the LCS PM reading between the background site (UBS) and the traffic site (TRS) is mainly due to the former having less particles with diameter < 600 nm compared to the traffic site and the inability of the Vaisala unit to detect below this cutoff. This conclusion cannot explain why the LCS UBS is still significantly biased high compared to the TRS during dust episodes (Fig. 4e) when the PM is expected to be dominated by large diameter sized particles
Page 8 first paragraph “ …. PM10 concentrations, however, are generally higher at the TRS compared to the UBS, as indicated by the reference instruments (Fig. 3g), which is in contrast to what the LCS measurements indicate (Fig. 3h)” Figures 3g and 3h are scatter plots for CO and NO2 respectively that have no relation to PM10. Authors need to correct this.
Page 8 second paragraph “ …. while those measured by the reference instruments were below 6–10 µg/m3” this statement is wrong. The difference for the reference instruments is never negative between the TRS and USB site as shown in Fig. 4f (solid line).
Authors need to add legends to Fig 4, Fig 5, Fig S14 and Fig S15 to help the readers.
Suggest the authors annotate Fig S2 to show when the firmware was changed for the NO2 and O3 sensors.
Table 4 appears to be an “image” and the font size is too small making it difficult to read. I suggest presenting this as an actual table and to increase the font size.
Viewed
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 97 | 0 | 1 | 98 | 0 | 0 |
- HTML: 97
- PDF: 0
- XML: 1
- Total: 98
- BibTeX: 0
- EndNote: 0
Viewed (geographical distribution)
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1