Bias in PM_2.5 measurements using collocated reference-grade and optical instruments

Area of Science:

• Environmental Science
• Atmospheric Science
• Air Quality Monitoring

Background:

• Optical PM2.5 measurement devices are sensitive to varying aerosol properties.
• Reference-grade instruments like beta attenuation monitors (BAMs) provide a benchmark for comparison.
• Understanding discrepancies between optical and reference PM2.5 measurements is vital for accurate air quality assessment.

Purpose of the Study:

• To compare PM2.5 measurements from collocated optical instruments (DustTrak II, DRX) and reference-grade BAMs over six months.
• To evaluate inter-model, intra-model, and inter-type instrument performance under ambient conditions.
• To assess the effectiveness of humidity correction methods for optical PM2.5 monitors.

Main Methods:

• Collocated optical (DustTrak II, DustTrak DRX) and reference-grade (BAM) PM2.5 monitors were deployed for six months.
• Inter-model, intra-model, and inter-type comparisons were conducted.
• Normalized Root Square Difference (NRMSD) and Normalized Root Mean Square Error (NRMSE) were calculated.
• Two humidity correction methods were explored to address bias in optical measurements.

Main Results:

• Averaged PM2.5 concentrations: DustTrak II (46.0, 45.5 μg/m³), DRX (29.8, 38.4 μg/m³), BAMs (18.3, 19.0 μg/m³).
• Intra-model precision: DustTrak II (~5% NRMSD), DRX (~27% NRMSD), BAMs (~15% NRMSD).
• Optical instruments showed significant bias relative to BAMs (NRMSE ~165% for DustTrak II, ~74% for DRX without correction).
• Humidity correction improved DRX NRMSE to ~27% when combined with BAM correction.
• Different correction methods yielded results differing by ~50%.

Conclusions:

• Optical PM2.5 instruments demonstrate high precision but exhibit substantial bias compared to reference-grade BAMs.
• Literature-derived calibration equations may not be universally applicable, necessitating local, model-specific calibration.
• This study provides valuable insights into the performance of optical PM2.5 monitors and the challenges in achieving accurate air quality data.