Performance evaluation and cross-validation of low-cost particulate matter sensors for environmental research
Loading...
Date
2025-12-10
Journal Title
Journal ISSN
Volume Title
Publisher
Plovdiv University Press "Paisii Hilendarski"
Abstract
Particulate monitoring data plays a vital role in supporting analysis, policymaking, and citizen initiatives, especially in areas related to ecology, air quality, public health, and overall quality of life. However, traditional regulatory monitoring systems are expensive and have drawbacks: they do not provide real-time data and cover only a limited number of official locations. As an alternative, low-cost monitoring devices are increasingly being used, but concerns remain regarding their accuracy and the reliability of the data they produce. As part of efforts to validate low-cost monitoring approaches, this work presents the design, implementation, and ongoing development of a low-cost, sensor-based air quality monitoring system dedicated to monitoring fine particulate matter (PM) and other atmospheric indicators within the METER.AC network. The system integrates multiple devices with various sensors, such as Honeywell HPMA115S0, Sensirion SEN55, and related SEN series, along with GPS modules for precise timestamp and geolocation. Data acquisition is synchronized and automated via UNIX shell scripts, which extract, convert, and process measurement data into structured CSV files containing parameters like PM1, PM2.5, PM10, temperature, humidity, etc. Fluke 985 Particle Counter is a high-quality professional device used as an independent benchmark for crossvalidation, providing particle counts for sizes: 0.3, 0.5, 1.0, 2.0, 5.0, and 10.0 µm. Measurement data are collected at 10-minute intervals and uploaded to a public visualization platform, where interactive graphs and summaries are generated. Also, this study aims to approximate the mass of airborne particulate matter based on particle size distribution derived from the FLUKE device output. On the basis of particle counts, the underlying particle size distribution is estimated with regression models.
Description
Keywords
fine particles, monitoring network, meter.ac, optical particle counter, MLRA