Air Quality Sensors: What They Measure and How to Interpret the Data

Which Parameters Sensors May Measure

Common devices measure or estimate fine particulate matter (PM2.5 and PM10), carbon dioxide, selected gaseous pollutants, volatile organic compounds, temperature, and relative humidity. A distinction is needed between direct measurement and model-based estimation. An optical particle sensor detects light scattered by aerosols, and electronics convert the signal into an estimated mass concentration. The result depends on particle size, shape, composition, and humidity — it is not the same as a reference gravimetric measurement.

PM2.5, PM10, CO₂, and Gases

Optical particle sensors are useful for showing short-term changes, peaks, and differences between locations, and can reveal possible effects from traffic, heating, dusty work, fires, or ventilation. However, smoke, fog, high humidity, and unusual aerosol composition can alter sensor response. Different models may not be directly comparable without adjustment. Indoors, CO₂ is often used as a ventilation indicator, but a normal CO₂ reading does not exclude particles, carbon monoxide, formaldehyde, or other substances. Total-VOC sensors generally respond to an overall change in the gas mixture and do not identify exact composition without specialised analysis. Electrochemical gas sensors may respond to interfering gases and should be evaluated under real operating conditions.

Consumer, Community, and Professional Systems

Consumer devices prioritise simplicity and rapid feedback. Community networks may combine many low-cost nodes to reveal spatial patterns. Professional and regulatory stations use standardised methods, quality systems, traceable calibration, and documented maintenance. A low-cost sensor can supplement official monitoring but does not replace a reference method where legally defensible evidence, compliance assessment, or precise chemical identification is required.

Sensor Placement and Representativeness

Placement determines which air mass a sensor represents. A device beside a road, kitchen, ventilation outlet, open window, or dust source will describe a local influence rather than the average condition of a neighbourhood. Comparable data require consistent height, distance from obstacles and sources, protection from rain and direct sunlight, stable power, and documented coordinates. Moving a sensor changes the context of the time series and should be recorded.

Calibration, Co-location, and Drift

Calibration links sensor response to a known value or reference instrument. A practical approach is to co-locate the sensor with a reliable station under similar conditions and estimate bias, precision, and sensitivity to weather. Performance may change because of ageing, contamination, condensation, or component replacement. Repeated checks, comparison among neighbouring devices, maintenance logs, and flags for unreliable periods are essential.

Interpreting a Time Series and Acting on Data

Examine the trend, episode duration, recurrence, background level, and relationship with weather or activity — not just a single reading. A minute-scale spike may be a genuine local event, a technical fault, or a brief humidity effect. Before conclusions, compare several sensors, an official station, meteorological information, and known sources. Hourly or daily aggregation may be needed when comparing data with guidance values. A monitoring network should start with a clear question: locating a source, comparing zones, assessing ventilation, tracking an incident, or evaluating an intervention. Where readings suggest a possible hazard, decisions should not rely on one low-cost device alone; professional measurement by a competent organisation may be required.