Can low-cost sensors (LCS) enhance air quality monitoring for personal pollution exposure assessment?

Abstract Laboratory and field assessments of low-cost sensors (LCS) are essential for ensuring the accuracy of PM 2.5 measurements collected by citizens in air quality campaigns. Evaluation of Sensirion SPS30 (LCS SPS30) in controlled laboratory setting showed a coefficient of determination ( R 2 ) ranging from 0.81–0.99 and a root mean square error (RMSE) from 0.81–61.72 μg m −3 , at average concentration of 21.5 μg m −3 . In contrast, co-location assessment at an average concentration of 9 μg m −3 resulted in R 2 of 0.5 and a RMSE of 6.82 μg m −3 . The results demonstrated that the sensor met micro-environmental monitoring standards (accuracy < 25%) and United States Environmental Protection Agency's performance criteria (RMSE ≤ 7 μg m −3 , R 2 > 0.7) only at relative humidity (RH) levels below 60%, emphasising its strong sensitivity to RH and the need for RH-dependent data corrections. The observed underestimation or overestimation of PM 2.5 readings was primarily attributed to variations in particle composition and concentration. Despite accuracy variations, LCSs can effectively capture spatiotemporal urban air quality patterns and identify pollution hotspots in community monitoring, particularly in low-pollution environments. In a citizen-led PM 2.5 monitoring campaign in Maribor, Slovenia, the lowest concentrations were recorded at 15:00 (2.9 μg m −3 ), while the highest occurred during the morning rush-hour (4.8 μg m −3 ), likely attributed to the planetary boundary layer’s impact on atmospheric particulate dispersion. Spatial analysis revealed that hotspots clustered near intersections, where vehicle waiting time is the longest.