Module 4: Data interpretation and accuracy (Open Source Air Quality Monitoring Toolkit)

Abstract

Low-cost air quality sensors do not automatically give perfectly accurate readings. Calibration is the process of adjusting sensor measurements to bring them closer to accurate values. Without it, raw readings may not correctly represent actual pollution levels. Once the network is deployed, data quality also needs to be maintained over time, because sensors can develop faults or their readings can gradually shift away from accurate values even when they appear to be working normally.

Knowing that the readings are accurate is only one part of the work. The other part is interpreting what they actually show. Patterns appear in air quality data regularly, but a visible pattern does not automatically prove a cause. Describing a pattern responsibly requires context, and claiming more than the evidence supports can undermine the credibility of the project's findings.

Navigating This Module

Read this module after the network has been defined and before results from your
monitors are shared externally. It remains relevant throughout the project's life, since calibration and quality checks do not end at deployment. Return to it whenever monitors are moved, reference data change, or unusual values appear in the readings. Work through the sections in order. First check how the system was calibrated and whether the monitors remain consistent. Then review routine quality checks, visualization choices, and interpretation rules before comparing the data to standards or guidelines.

Access the full toolkit here.