Indication Variability of the Particulate Matter Sensors Dependent on Their Location

Area of Science:

• Environmental Science
• Public Health
• Sensor Technology

Background:

• Air quality, particularly during the heating season, is a significant concern in Poland, with rural areas often experiencing worse conditions due to domestic coal burning.
• Anthropogenic particulate matter (PM) poses substantial health risks, especially in regions with limited official air quality monitoring.

Purpose of the Study:

• To assess the impact of sensor placement on particulate matter (PM) readings within a localized air quality monitoring network.
• To evaluate the effectiveness of a distributed sensor network for representative air quality assessment and anomaly detection.

Main Methods:

• Deployment of a network of low-cost PMS7003 PM sensors and ESP8266 microcontrollers with Wi-Fi.
• Analysis of PM sensor data, comparing readings from geographically close sensors under various weather and pollution conditions.
• Development of a method to detect sudden air quality deterioration using the difference between mean and median sensor readings.

Main Results:

• Particulate matter sensor readings can vary by up to tenfold over short distances (dozens of meters), influenced by weather and pollution sources.
• A distributed sensor network provides more representative air quality measurements than single-point monitoring.
• A method comparing mean and median sensor values effectively detects sudden increases in air pollutants, with differences up to 400 µg/m³ over 5 minutes.

Conclusions:

• Distributed, low-cost sensor networks offer a more accurate representation of air quality, especially in underserved areas.
• The developed anomaly detection method enables rapid identification of air quality degradation, facilitating timely public health interventions.
• This approach is valuable for protection systems requiring swift detection of environmental anomalies.