Journal metrics

Journal metrics

  • IF value: 3.248 IF 3.248
  • IF 5-year value: 3.650 IF 5-year 3.650
  • CiteScore value: 3.37 CiteScore 3.37
  • SNIP value: 1.253 SNIP 1.253
  • SJR value: 1.869 SJR 1.869
  • IPP value: 3.29 IPP 3.29
  • h5-index value: 47 h5-index 47
  • Scimago H index value: 60 Scimago H index 60
Discussion papers | Copyright
https://doi.org/10.5194/amt-2018-285
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

Research article 10 Sep 2018

Research article | 10 Sep 2018

Review status
This discussion paper is a preprint. It is a manuscript under review for the journal Atmospheric Measurement Techniques (AMT).

An improved low power measurement of ambient NO2 and O3 combining electrochemical sensor clusters and machine learning

Kate R. Smith1, Peter M. Edwards1, Peter D. Ivatt1, James D. Lee1,2, Freya Squires1, Chengliang Dai1, Richard E. Peltier3, Mat J. Evans1,2, and Alastair C. Lewis1,2 Kate R. Smith et al.
  • 1Wolfson Atmospheric Chemistry Laboratories, University of York, York, YO10 5DD, United Kingdom
  • 2National Centre for Atmospheric Science, University of York, York YO10 5DD, United Kingdom
  • 3Environmental Health Science, University of Massachusetts, 686 North Pleasant Street Amherst, MA 01003, USA

Abstract. Low cost sensors (LCS) are an appealing solution to the problem of spatial resolution in air quality measurement, but they currently do not have the same analytical performance as regulatory reference methods. Individual sensors can be susceptible to analytical cross interferences, have random signal variability and experience drift over short, medium and long timescales. To overcome some of the performance limitations of individual sensors we use a clustering approach using the instantaneous median signal from six identical electrochemical sensors to minimise the randomised drifts and inter-sensor differences. We report here a low power analytical device (<200W) that comprises of clusters of sensors for NO2, OX, CO and total VOC, and that measures supporting parameters such as water vapour and temperature. This was tested in the field against reference monitors, collecting ambient air pollution data in Beijing, China. Comparisons were made of NO2 and OX clustered sensor data against reference methods for calibrations derived from factory settings, in-field simple linear regression (SLR) and then against three machine learning (ML) algorithms. The parametric supervised ML algorithms boosted regression trees (BRT) and boosted linear regression (BLR) and the non-parametric technique Gaussian Process (GP) used all available sensor data to improve the measurement estimate of NO2 and OX. In all cases ML produced an observational value that was closer to reference measurements than SLR alone. In combination, sensor clustering and ML generated sensor data of a quality that was close to that of regulatory measurements (using the RSME metric) yet retained a very substantial cost and power advantage.

Download & links
Kate R. Smith et al.
Interactive discussion
Status: open (until 05 Nov 2018)
Status: open (until 05 Nov 2018)
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
[Subscribe to comment alert] Printer-friendly Version - Printer-friendly version Supplement - Supplement
Kate R. Smith et al.
Data sets

Electrochemical sensor data K. R. Smith, P. M. Edwards, and A. C. Lewis https://doi.org/10.15124/1a0c64b0-433b-4eec-b5c7-64d3de0a0351

Kate R. Smith et al.
Viewed
Total article views: 225 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
182 40 3 225 1 4
  • HTML: 182
  • PDF: 40
  • XML: 3
  • Total: 225
  • BibTeX: 1
  • EndNote: 4
Views and downloads (calculated since 10 Sep 2018)
Cumulative views and downloads (calculated since 10 Sep 2018)
Viewed (geographical distribution)
Total article views: 225 (including HTML, PDF, and XML) Thereof 224 with geography defined and 1 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Cited
Saved
No saved metrics found.
Discussed
No discussed metrics found.
Latest update: 18 Sep 2018
Publications Copernicus
Special issue
Download
Short summary
Clusters of low- cost, low-power atmospheric gas sensors were built into a sensor instrument to monitor NO2 and O3 in Beijing, alongside reference instruments, with an aim to improve the reliability of sensor measurements. Clustering identical sensors and using the median sensor signal was used to minimise drift over short and medium timescales. Three different machine learning techniques used all the sensor data in an attempt to correct for cross interferences, which worked to some degree.
Clusters of low- cost, low-power atmospheric gas sensors were built into a sensor instrument to...
Citation
Share