Calibration of low-cost particulate matter sensors: Model development for a multi-city epidemiological study

Abstract

Low-cost air monitoring sensors are an appealing tool for assessing pollutants in environmental studies. Portable low-cost sensors hold promise to expand temporal and spatial coverage of air quality information. However, researchers have reported challenges in these sensors' operational quality. We evaluated the performance characteristics of two widely used sensors, the Plantower PMS A003 and Shinyei PPD42NS, for measuring fine particulate matter compared to reference methods, and developed regional calibration models for the Los Angeles, Chicago, New York, Baltimore, Minneapolis-St. Paul, Winston-Salem and Seattle metropolitan areas. Duplicate Plantower PMS A003 sensors demonstrated a high level of precision (averaged Pearson's r = 0.99), and compared with regulatory instruments, showed good accuracy (cross-validated R² = 0.96, RMSE = 1.15 µg/m³ for daily averaged PM_2.5 estimates in the Seattle region). Shinyei PPD42NS sensor results had lower precision (Pearson's r = 0.84) and accuracy (cross-validated R² = 0.40, RMSE = 4.49 µg/m³). Region-specific Plantower PMS A003 models, calibrated with regulatory instruments and adjusted for temperature and relative humidity, demonstrated acceptable performance metrics for daily average measurements in the other six regions (R² = 0.74-0.95, RMSE = 2.46-0.84 µg/m³). Applying the Seattle model to the other regions resulted in decreased performance (R² = 0.67-0.84, RMSE = 3.41-1.67 µg/m³), likely due to differences in meteorological conditions and particle sources. We describean approach to metropolitan region-specific calibration models for low-cost sensors that can be used with cautionfor exposure measurement in epidemiological studies.

Keywords: Air pollution; Air quality system network (AQS); Calibration; Fine particulate matter; Low-cost monitors (LCM); Multicenter study design.

Figure 1:

Map of regions covered by the ACT-AP and MESA Air studies with the locations of PM_2.5 reference sites.

Figure 2a:

Correlation between duplicate sensors within a box Note: Plantower PMS A003 measures are the manufacturer calculated daily averaged PM_2.5 mass; Shinyei PPD₄₂NS measures are daily averaged raw sensor readings (low pulse occupancy). Data from all regions are included.

Figure 2b:

Relationship of Plantower PMS A003 and Shinyei PPD₄₂NS PM_2.5 daily scale model predictions with PM_2.5 FRM measurements in the Seattle metropolitan areas. Note: Both models were evaluated against the subset of 1308 FRM monitor-days in Seattle with both Plantower PMS A003 and Shinyei PPD₄₂NS data.

Figure 3:

The comparison of distributions of calibration model residuals between different types of models in each region. Note: Figure shows boxplots of model residuals from the FRM daily model (10-fold cross-validation), FRM & FEM daily model (10-fold cross-validation), and Seattle out-of-region model. The models were evaluated against FRM PM_2.5 monitoring sites;

Figure 4:

Correlation between daily averaged temperature (°F) and RH measures within low-cost monitors across different regions Note: low-cost sensors readings of temperature and RH were calibrated with reference temperature/RH data from Seattle, Beacon Hill site in order to present standard units.

Figure 5:

Residuals by temperature (°F) from different calibration models in LA and Chicago. Note: low-cost sensors readings of temperature were calibrated with reference temperature data from Seattle, Beacon Hill site in order to present standard units.