Systematic studies of commodity particulate matter air pollution sensors

Air pollution exposure is a leading risk factor for morbidity and mortality, with particulate matter (PM) generally being associated with the greatest health impact of all measured pollutants. PM concentrations vary spatiotemporally from micro- to cityto regional-scales, depending on sources and meteorological conditions. Commodity PM sensors, defined as PM sensors with characteristics that make them suitable for large scale deployment (size, ease of use, price and frequency of measurement), often qualified as ’low-cost’, are based on nephelometry and they could increase the spatiotemporal resolution of current monitoring. However, questions remain about their performance variation, including with PM characteristics and meteorological factors. There is no current agreed best practice to certify the performances of these sensors, especially at fine temporal resolutions. This study aimed to characterise multiple commodity PM sensors by comparing their performance to reference-grade instruments and to evaluate methods to calibrate them. The initial work developed the capabilities to utilise multiple commodity PM sensors in outdoor studies. Next, the performances of multiple commodity PM sensor models were compared during a yearlong outdoor campaign. To further characterise observed behaviour, a laboratory study was then conducted focusing on the transient response of the sensors. Finally, the sensors were collocated with a reference instrument with a similar time resolution, to evaluate a range of calibration models and strategies and propose improvements to existing sensor certifications. Sensors from different manufacturers demonstrated differential response to relative humidity. Sensors could detect transient PM peaks in the field and under controlled conditions. In the latter, a differential response was demonstrated for two combustion sources of PM, varying by sensor model and humidity. For calibration at a 2 min resolution, models incorporating sensor-reported particle number concentration, coupled with pre- and post-deployment calibration, showed potential for correcting these sensors to reach uncertainty around 25% and precision <2.5 µg/m3, similar to reference instruments minimum standards. These results show that, with careful calibration and assessment of environmental conditions, commodity PM sensors can accurately report PM air pollution concentrations and that there is scope to improve existing certification efforts. Further work is needed to evaluate their applicability in specific use-case scenarios, such as assessment of personal exposure while moving within different environments, or assessment of the impact of mitigation measures feeding back to policymaking or facilitating source identification