Long-term stability of low-cost air quality sensors for NO and NO₂

The study investigated the long-term stability of low-cost air quality sensors, focusing on their calibration through co-location with the Swiss national monitoring network. Robust linear regression and random forest regression models were employed to calibrate the sensors. The study analyzed sensor performance during deployment in the city of Zurich and re-evaluated them upon return to the original site. The results highlighted the reliability of these sensors over time, revealed that NO₂ sensors exhibited greater degradation compared to NO sensors, especially after relocation, underscoring the need for careful recalibration in long-term applications. Furthermore, the findings demonstrated that NO₂ sensors were prone to significant biases, particularly in low-pollution regions, which limits their use without frequent recalibration.

Check out Master’s Project Work and Journal Paper for more information.