New Research Shows Communities Need Better Guidance on Wildfire Air Quality

EMBARGOED UNTIL August 20, 2026 @ 8:00 a.m.

Study finds low-cost air sensors help residents protect themselves, but disparities persist in who the information they produce

August 20, 2026 (Herndon, VA) — During one of Oregon’s most destructive wildfire seasons on record in 2020, 1,200 residents surveyed in the weeks after the fires reported being overwhelmed by air quality information. They received data from multiple sources but little clear guidance on how to respond. A new study, published in Risk Analysis, finds that low-cost internet-connected air sensors can change that by providing hyper-local, real-time information that helps residents take protective action. But not all communities have equal access to this technology.

Traditional air quality monitoring relies on sensors operated by government agencies, designed to track industrial pollution. These sensors report data averaged over 6- to 12-hour periods and are often delayed. In the age of fast-moving wildfires, this infrastructure leaves residents making protective decisions with incomplete information. Private individuals and organizations can now deploy internet-connected particulate matter low-cost sensors for $250. Thousands of these devices are currently connected to public sites like AirNow.gov, creating unprecedented hyper-local air quality data. Yet despite their rapid growth, existing research has not looked at who actually uses this information or whether it changes protective behavior.

Researchers at the University of Oregon partnered with the Oregon Health Authority to survey 1,200 residents, including rural and Spanish-speaking communities, who had experienced the 2020 wildfires. They applied the Protective Action Decision Model to understand how people process hazard information and decide whether to protect themselves. 

Key Findings

• Sensor data changes protective behavior: Residents who accessed air quality data from low-cost sensors were significantly more likely to take protective actions, such as using air filters or wearing N95 masks indoors. People who used sensor data had more than twice the odds of filtering indoor air compared with those who did not. For indoor masking, the effect was similarly strong. This link held across different levels of smoke exposure, suggesting that awareness of air quality drives action independent of how much smoke residents actually breathed.
• Only 28% of residents used publicly available sensor data: Despite the proliferation of these sensors, more than 72% of respondents from the fire-affected areas of Oregon had never accessed the most popular information sites.  Users tended to be younger, living in cities and well-educated. The number of nearby sensors did not predict whether people used them, indicating that sensor availability alone does not drive use. The research suggests that awareness, accessibility and the ability to interpret data matter more than the physical presence of sensors nearby.
• Urban residents took more protective actions than rural residents: Across three of the four protective behaviors measured, rural residents had significantly lower odds of acting. Rural residents were 37% less likely to wear masks indoors and 40% less likely to wear masks outdoors than urban residents. 
• Financial insecurity constrained protective action: Households that reported rarely having enough money to pay bills showed significantly lower odds of using air filtration systems and wearing masks indoors. Those who reported sometimes having enough money were 86% less likely to use filtration than those who always had sufficient income. Access to protective equipment requires resources, and cost barriers left lower-income residents more exposed to smoke risks.
• People with disabilities took more protective actions despite facing barriers: Residents reporting disabilities had substantially higher odds of evacuating due to smoke and of wearing masks indoors and outdoors. Yet disability status was not significantly associated with the use of air filtration systems, likely due to cost and accessibility barriers. These findings suggest that people with disabilities perceive themselves as more vulnerable to smoke, but economic constraints limit their access to protective measures.

“There remain systematic barriers to producing air quality data individually, as the costs are quite high for a single-purpose device,” said Benjamin Y. Clark, professor at the University of Oregon and an author of the study. “Increasing the availability of air data can lead to more protective actions. Governments and nonprofits have the potential to enhance the visibility and accessibility of sensor data, therefore increasing engagement.”

Authors Available for Interview

Benjamin Y. Clark, corresponding author, School of Planning, Public Policy and Management, University of Oregon

Shahinur Bashar, School of Planning, Public Policy and Management, University of Oregon

“How IoT Enables Protective Action Decision Making During a Wildfire Event,” Risk Analysis, 2026

###

About Risk Analysis

Risk Analysis is a peer-reviewed journal publishing original research on the assessment and management of risks across disciplines including public health, engineering, environmental science, social science and policy. Founded in 1980, Risk Analysis is published by Wiley on behalf of the Society for Risk Analysis.

Media Contact

Anya Grondalski, Big Voice Communications
anya@bigvoicecomm.com