Enhancing PM_2.5 Measurement Accuracy: Insights from Environmental Factors and BAM-Light Scattering Device Correlation

Table 1

Literature on errors according to environmental factors of low-cost light scattering device.

Literature	Comparison device	Environmental factor	Test condition	Analysis method	Influential factor
Wu et al. [22]	BAM; TEOM	(i) Temperature (-10–50°C); (ii) Relative humidity (20–95%)	Lab/field	Statistic	(i) Particle size (1.1, 2.0, 2.5, 3.0, and 8.0 μm, PM1.0, PM2.5, and PM10)
Molnár et al. [23]	BAM	(i) Humidity (40–100%)	Field	Statistic	—
Tryner et al. [24]	TEOM; SMPS; APS	(i) Humidity (15–90%)	Lab	(i) Statistic; (ii) PyMieScatt (Python)	(i) Contaminated sensor (PM2.5: measured over 7300 μg/m^3, PM10: 33,000 μg/m^3); (ii) PM type (ammonium sulfate, Arizona road dust, National Institute of Standards and Technology (NIST) Urban PM, and wood smoke)
Han et al. [25]	ELPI	(i) Humidity (5–80%)	Lab	Statistic	(i) PM type (fly ash and pure mineral)
Levy Zamora et al. [26]	BAM	(i) Humidity (20–80%)	Lab/field	Statistic	(i) PM type (incense, oleic acid, NaCl, talcum powder, cooking emissions, and monodispersed polystyrene latex spheres)
Olivares and Edwards [27]	TEOM	(i) Temperature (6–26°C)	Lab	(i) Statistic; (ii) Openair R package (R Team)	(i) Concentration (0–170 μg/m^3)
Present study	BAM; LS	(i) Temperature (11–31°C); (ii) Humidity (39–96%); (iii) Pressure (989–1008 hPa); (iv) Precipitation (0.0–0.4 mm); (v) Temperature–dew point temperature (0.6–15.4°C); (vi) Wind speed (0.0–4.9 m/s); (vii) Wind direction (16 directions)	Field	(i) Statistic; (ii) Machine learning (Orange)	(i) Ambient environment; (ii) Omnidirectional inlet design of device (to detect wind direction in all directions)