Assessing low-cost sensor performance at varying temporal resolution against reference instruments for the measurement of NO2

Air pollution continues to be a major global health concern, with nitrogen dioxide (NO2) significantly contributing to negative health impacts. Low-cost sensors (LCS) present promising opportunities for accessible, high-resolution air quality monitoring but are often questioned for their accuracy and reliability. This study assesses the performance of electrochemical LCS for NO₂ measurements compared to high-precision reference instruments—cavity attenuated phase shift (CAPS) and chemiluminescence NO₂ monitors—across eleven temporal resolutions (ranging from 10 seconds to 6 hours). Data were collected over six months at an urban-traffic air quality monitoring site in Berlin using three EarthSense Zephyr sensor systems equipped with electrochemical sensors. Statistical metrics, including R², relative error (%), and mean bias error (MBE), were used to evaluate sensor performance. The results indicate that LCS demonstrate strong agreement with reference instruments at coarse time resolutions (≥1-hour averages, R² > 0.8), but their accuracy declines considerably at higher resolutions (<1-minute, R² < 0.5). Performance improves when sensors are calibrated against CAPS monitors compared to chemiluminescence monitors. Factors such as chemistry and emissions play a significant role, with poorer performance during the day than at night, a discrepancy that is further amplified at finer temporal resolutions. CAPS-calibrated predictive models also excel in capturing short-term concentration peaks compared to those calibrated with chemiluminescence monitors. These findings highlight that while LCS are effective for coarse-resolution NO₂ monitoring, their limitations in dynamic environments at high temporal resolutions pose challenges for use in exposure studies and mobile applications. The study recommends careful calibration, strategic experimental design, and a focus on lower time-resolution applications to enhance LCS deployment.