Intra-urban spatial variability of surface ozone in Riverside, CA: viability and validation of low-cost sensors

Kira Sadighi, Evan Coffey, Andrea Polidori, Brandon Feenstra, Qin Lv, Daven K. Henze, Michael HanniganAtmospheric 2018 · DOI

Abstract

Sensor networks are being more widely used to characterize and understand compounds in the atmosphere like ozone ($O_3$). This study employs a measurement tool, called the U-Pod, constructed at the University of Colorado Boulder, to investigate spatial and temporal variability of $O_3$ in a 200 $km^2$ area of Riverside County near Los Angeles, California. This tool contains low-cost sensors to collect ambient data at non-permanent locations. The U-Pods were calibrated using a pre-deployment field calibration technique; all the U-Pods were collocated with regulatory monitors. After collocation, the U-Pods were deployed in the area mentioned. A subset of pods was deployed at two local regulatory air quality monitoring stations providing validation for the collocation calibration method. Field validation of sensor $O_3$ measurements to minute-resolution reference observations resulted in $R^2$ and root mean squared errors (RMSEs) of 0.95–0.97 and 4.4–5.9 ppbv, respectively. Using the deployment data, ozone concentrations were observed to vary on this small spatial scale. In the analysis based on hourly binned data, the median $R^2$ values between all possible U-Pod pairs varied from 0.52 to 0.86 for ozone during the deployment. The medians of absolute differences were calculated between all possible pod pairs, 21 pairs total. The median values of those median absolute differences for each hour of the day varied between 2.2 and 9.3 ppbv for the ozone deployment. Since median differences between U-Pod concentrations during deployment are larger than the respective root mean square error values, we can conclude that there is spatial variability in this criteria pollutant across the study area. This is important because it means that citizens may be exposed to more, or less, ozone than they would assume based on current regulatory monitoring.

Introduction

Regulatory air quality monitoring stations (AQMS) provide the "gold standard" for atmospheric data but are limited by high capital and maintenance costs, often exceeding $100,000 per station. Consequently, these stations are sparse, often leaving gaps of 10 kilometers or more in urban environments. Ozone ($O_3$) is a complex secondary pollutant formed by reactions between nitrogen oxides ($NO_x$) and volatile organic compounds (VOCs). Because these precursors originate from localized sources like traffic and industry, ozone levels can fluctuate significantly within a single city. This study investigates whether low-cost sensor networks can reliably capture this "intra-urban" variability, providing a more granular view of public exposure than sparse regulatory networks allow.

Methods

The research team utilized the U-Pod, a portable sensing platform developed at the University of Colorado Boulder. Each U-Pod integrates a MiCS 2611 metal oxide ($MO_x$) sensor for ozone, along with sensors for temperature, humidity, and GPS tracking.

Internal components (a) and field deployment setup (b) of the U-Pod sensing platform.

To ensure data accuracy, the researchers employed a "field calibration" strategy rather than laboratory calibration. Thirteen U-Pods were collocated with a high-grade regulatory monitor for three weeks prior to deployment. A "Linear 4T" regression model was developed to convert raw sensor resistance into ozone concentrations, specifically accounting for temperature, absolute humidity, and temporal drift. Following calibration, the pods were deployed across a 200 $km^2$ area in Riverside County, covering diverse industrial, residential, and commercial zones.

Key Results and Discussion

The study successfully validated the U-Pod's performance, with calibration phases yielding $R^2$ values between 0.97 and 0.99. During the 10-week field deployment, validation against reference monitors resulted in a root mean squared error (RMSE) of 4.4–5.9 ppbv. Given that regulatory limits are typically around 70 ppbv, this error margin is acceptable for spatial mapping applications.

Crucially, the data confirmed real spatial divergence in ozone levels across the study area. Median $R^2$ values between different U-Pod pairs varied from 0.52 to 0.86; if the air were uniform, these values would approach 1.0. Furthermore, the median difference in ozone between sites (up to 9.3 ppbv) was consistently larger than the sensor's error margin (~5 ppbv), proving that the network was detecting actual neighborhood-scale differences rather than random noise. The researchers also noted a "claw-like" pattern in some data, likely caused by localized ozone titration from fresh $NO_x$ emissions (e.g., heavy truck traffic) near specific sensor sites.

Conclusion and Future Work

This study demonstrates that low-cost $MO_x$ sensors are viable tools for characterizing intra-urban ozone variability when supported by robust field calibration. While the sensors showed remarkable stability over a three-month period, the authors acknowledge that future iterations should include $NO_x$ sensors to account for cross-sensitivity and further refine the ozone models. Additionally, reducing the logistical burden of long collocation periods remains a key area for future research to make these networks more scalable for urban air quality management.

Summary updated

Abstract

Sensor networks are being more widely used to characterize and understand compounds in the atmosphere like ozone (O3). This study employs a measurement tool, called the U-Pod, constructed at the University of Colorado Boulder, to investigate spatial and temporal variability of O3 in a 200 km2 area of Riverside County near Los Angeles, California. This tool contains low-cost sensors to collect ambient data at non-permanent locations. The U-Pods were calibrated using a pre-deployment field calibration technique; all the U-Pods were collocated with regulatory monitors. After collocation, the U-Pods were deployed in the area mentioned. A subset of pods was deployed at two local regulatory air quality monitoring stations providing validation for the collocation calibration method. Field validation of sensor O3 measurements to minute-resolution reference observations resulted in R2 and root mean squared errors (RMSEs) of 0.95–0.97 and 4.4–5.9 ppbv, respectively. Using the deployment data, ozone concentrations were observed to vary on this small spatial scale. In the analysis based on hourly binned data, the median R2 values between all possible U-Pod pairs varied from 0.52 to 0.86 for ozone during the deployment. The medians of absolute differences were calculated between all possible pod pairs, 21 pairs total. The median values of those median absolute differences for each hour of the day varied between 2.2 and 9.3 ppbv for the ozone deployment. Since median differences between U-Pod concentrations during deployment are larger than the respective root mean square error values, we can conclude that there is spatial variability in this criteria pollutant across the study area. This is important because it means that citizens may be exposed to more, or less, ozone than they would assume based on current regulatory monitoring.
BibTeX
@inproceedings{sadighi2018intra,
  author = {Kira Sadighi and Evan Coffey and Andrea Polidori and Brandon Feenstra and Qin Lv and Daven K. Henze and Michael Hannigan},
  title = {{Intra-urban spatial variability of surface ozone in Riverside, CA: viability and validation of low-cost sensors}},
  booktitle = {Atmospheric},
  year = {2018},
  doi = {10.5194/amt-11-1777-2018},
}
Figure