Prolonged exposure to particulate matter (PM) is associated with adverse impacts on human health leading to pulmonary and cardiovascular diseases. 1-6 Rapid economic growth accompanied with a spur in industrialization, urbanization and energy consumption has led to increased emissions of PM which has put the health of citizens at stake. In 2012 alone, 0.6 million pre-mature deaths and loss of 25 million disability-adjusted life years could be attributed to ambient air pollution in India. 7
Conventionally, ambient PM monitoring is carried out at static and sparse research facilities or government environmental monitoring agencies. For instance, in India, Central Pollution Control Board controls 342 air quality monitoring stations, out of which only 44 stations provide real-time PM measurements. This is clearly too inadequate to capture spatio-temporal variations in PM, accurately estimate exposure for a population of 1.25 billion, identify pollution hot spots and devise efficient strategies to combat this menace.
Moreover, due to frequent media attention and growing public awareness, there is a surge in the demand for real-time air quality data by concerned individuals who wish to monitor and regulate their exposure to ambient pollutants.
High investment costs incurred during installation and maintenance of PM analyzers have hindered extensive coverage and wide spread availability of measurements.
In the last few years immense progress has been made in the development of portable low-cost sensors by small- and medium-sized enterprises for providing real time information on PM levels. 8, 9. Most of these aforementioned sensors detect particles via a light scattering method. 10 These sensors have garnered widespread attention because of their low cost and their ability to characterize PM concentrations at a high spatial and temporal resolution. The data from the sensors is readily available to the user.
Without a doubt, low cost PM sensors have many promising applications. However, before large volume of untested data is widely put forth in public domain, the instruments must be properly validated against certified methods and the results should be made aware to the consumer of the data.
Several studies in the last two years have evaluated the performance of few commercially available particle sensors in laboratory. Wang, et al. 11 assessed the performance of three low-cost PM sensors, Shinyei PPD42NS, Samyoung DSM501A and Sharp against US EPA certified methods under laboratory conditions and reported high dependence on the particle composition, particle size and relative humidity. Manikonda, et al. 12 also evaluated the performance of four low-cost PM sensors (Speck, Dylos, TSI AirAssure and UB AirSense) using cigarette smoke and Arizona test dust under standard relative humidity and temperature conditions and found adequate precision for monitoring PM exposure in indoor environments. Austin, et al. 13 evaluated the performance of Shinyei PPD42NS under laboratory conditions using monodisperse polystyrene spheres and found the sensor appropriate for low to medium concentrations of respirable particles (< 100 µgm-3).
Several studies have also evaluated the performance of few lost-cost PM sensors under ambient conditions 10, 14-18. Han, Symanski and Stock 18 evaluated the performance of DC 1700 PM sensor using a portable aerosol spectrometer working on the principle of light scattering (GRIMM-11R) as a reference in an urban residential area with average PM2.5 mass loading of 11.3 µg/m3 (0.2–318.7 µg/m3) and average PM10-2.5 mass loading of 4.8 µg/m3 (0–378.7 µg/m3). They reported good agreement (r2 = 0.7 for PM2.5 ; r2 = 0.4 for PM10-2.5) between the sensor and the reference analyzer when relative humidity was less than 60% and ambient mass concentration was less than 300 µg/m3 for both coarse and fine PM. Holstius, Pillarisetti, Smith and Seto 14 evaluated the performance of a custom-built platform that used a Shinyei PPD42NS using an analyzer working on the principle of β-attenuation (BAM-1020, Met One Instruments) as a reference at a regulatory monitoring site in California with low ambient PM2.5 mass concentration between 2 to 21 µg/m3. They were able to explain 72% of the variance observed in 24 hour PM2.5 data based on linear corrections. Jiao and co-workers tested 5 different types of low cost PM sensors using β-attenuation analyzer (Met One, BAM 1020) as a reference, at a site in southeastern US with low ambient PM2.5 of ~10 µg/m3 and only 3 sensors namely Dylos, Shinyei PPD60PV and Shinyei PPDD42NS had an r value more than 0.5 10.
There is a need to test the efficacy of such low cost sensors for regular ambient air quality monitoring in extremely polluted environments with high levels of PM.