Low-cost sensors as an alternative for long-term air quality monitoring.

Low-cost air quality sensors are increasingly being used in many applications; however, many of their performance characteristics have not been adequately investigated. This study was conducted over a period of 13 months using low-cost air quality monitors, each comprising two low-cost sensors, which were subjected to a wide range of pollution sources and concentrations, relative humidity and temperature at four locations in Australia and China. The aim of the study was to establish the performance characteristics of the two low-cost sensors (a Plantower PMS1003 for PM2.5 and an Alphasense CO–B4 for carbon monoxide, CO) and the KOALA monitor as a whole under various conditions. Parameters evaluated included the inter-variability between individual monitors, the accuracy of monitors in comparison with the reference instruments, the effect of temperature and RH on the performance of the monitors, the responses of the PM2.5 sensors to different types of aerosols, and the long-term stability of the PM2.5 and CO sensors. The monitors showed high inter-correlations (r > 0.91) for both PM2.5 and CO measurements. The monitor performance varied with location, with moderate to good correlations with reference instruments for PM2.5 (0.44< R2 < 0.91) and CO (0.37< R2 < 0.90). The monitors performed well at relative humidity < 75% and high temperature conditions; however, two monitors in Beijing failed at low temperatures, probably due to electronic board failure. The PM2.5 sensor was less sensitive to marine aerosols and fresh vehicle emissions than to mixed urban background emissions, aged traffic emissions and industrial emissions. The long-term stability of the PM2.5 and CO sensors was good, while CO relative errors were affected by both high and low temperatures. Overall, the KOALA monitors performed well in the environments in which they were operated and provided a valuable contribution to long-term air quality monitoring within the elucidated limitations.