Calibration Drift In Low-Cost PM and Sound Sensors Across Contrasting Indoor Environments

This dataset contains measurements from a custom low-cost sensor network deployed in two contrasting indoor environments to evaluate calibration drift in particulate matter (PM₁₀) and sound level sensors. Deployments include (1) an industrial painting environment and (2) a swine barn environment.

The dataset includes measurements from three co-located low-cost monitors and co-located reference instruments collected during scheduled reference sampling periods. Variables include low-cost sensor PM₁, PM₂.₅, and PM₁₀ mass concentrations (µg/m³), A-weighted sound levels (dBA), temperature (°C), relative humidity (%), and timestamps, along with reference PM and sound measurements where available. Raw data are provided for both reference and low-cost instruments in CSV format and as .RData files. Accompanying R scripts used in the paper’s analysis are also included.

These data support analyses of time-dependent calibration drift using linear mixed-effects models, comparison of drift magnitude across environments with different particulate loading, and evaluation of drift-aware calibration approaches.