The EuNetAir Joint Exercise focused on the evaluation and assessment of environmental gaseous, particulate matter (PM) and meteorological microsensors versus standard air quality reference methods through an experimental urban air quality monitoring campaign. This work presents the second part of the results, including evaluation of parameter dependencies, measurement uncertainty of sensors and the use of machine learning approaches to improve the abilities and limitations of sensors. The results confirm that the microsensor platforms, supported by post processing and data modelling tools, have considerable potential in new strategies for air quality control. In terms of pollutants, improved correlations were obtained between sensors and reference methods through calibration with machine learning techniques for CO (r2 = 0.13–0.83), NO2 (r2 = 0.24–0.93), O3 (r2 = 0.22–0.84), PM10 (r2 = 0.54–0.83), PM2.5 (r2 = 0.33–0.40) and SO2 (r2 = 0.49–0.84). Additionally, the analysis performed suggests the possibility of compliance with the data quality objectives (DQO) defined by the European Air Quality Directive (2008/50/EC) for indicative measurements.

Assessment of air quality microsensors versus reference methods: The EuNetAir Joint Exercise – Part II

S. De Vito;E. Esposito;M. Salvato;D. Suriano;V. Pfister;M. Prato;S. Dipinto;M. Penza
2018-01-01

Abstract

The EuNetAir Joint Exercise focused on the evaluation and assessment of environmental gaseous, particulate matter (PM) and meteorological microsensors versus standard air quality reference methods through an experimental urban air quality monitoring campaign. This work presents the second part of the results, including evaluation of parameter dependencies, measurement uncertainty of sensors and the use of machine learning approaches to improve the abilities and limitations of sensors. The results confirm that the microsensor platforms, supported by post processing and data modelling tools, have considerable potential in new strategies for air quality control. In terms of pollutants, improved correlations were obtained between sensors and reference methods through calibration with machine learning techniques for CO (r2 = 0.13–0.83), NO2 (r2 = 0.24–0.93), O3 (r2 = 0.22–0.84), PM10 (r2 = 0.54–0.83), PM2.5 (r2 = 0.33–0.40) and SO2 (r2 = 0.49–0.84). Additionally, the analysis performed suggests the possibility of compliance with the data quality objectives (DQO) defined by the European Air Quality Directive (2008/50/EC) for indicative measurements.
2018
Experimental campaign
Low-cost microsensors
Machine learning
Measurement uncertainty
Air quality monitoring
Reference methods
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12079/60045
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
social impact