Skip to main content
TerraFuture
Research6 min read

Breathing Easier: Expanding Our Community Air Quality Sensor Network

Oregon DEQ operates 6 air quality monitors for all of Portland. Our network of 84 low-cost sensors reveals pollution variations of up to 340% within a single neighborhood — disparities that regulatory monitors cannot detect.

MC
Marcus Chen
Director of Research · May 20, 2025
Small air quality sensor mounted on a pole in an urban neighborhood with residential buildings in background

Air quality in Portland is typically described using data from the Oregon Department of Environmental Quality's regulatory monitoring network, which operates 6 stations across the metropolitan area. These stations produce high-quality, reference-grade data that serves its regulatory purpose well. But 6 data points across a city of 650,000 people and 145 square miles cannot capture the hyperlocal pollution patterns that determine actual exposure for individual communities.

TerraFuture launched its Community Air Quality Sensor Network in 2022 with 24 low-cost monitors. Today, the network has expanded to 84 sensors deployed across all four quadrants of Portland, producing continuous hourly readings that reveal air pollution patterns invisible to the regulatory network.

The Technology

Our network uses PurpleAir PA-II sensors, which measure PM2.5 particulate matter using laser particle counting technology. These sensors cost approximately 250 dollars each, compared to 15,000 to 250,000 dollars for federal reference method monitors. The tradeoff is precision: PurpleAir sensors have a measurement uncertainty of approximately 10 to 15 percent compared to 2 to 5 percent for reference instruments.

To improve data quality, every TerraFuture sensor undergoes a two-week colocation calibration period alongside the nearest DEQ reference monitor before deployment. We apply location-specific correction factors developed by the EPA and the University of Washington's MakerSpace program. Post-correction, our sensors agree with reference monitors to within 3 micrograms per cubic meter for 95 percent of hourly readings, which is sufficient for community health assessment.

All sensors report data in real time to our open data portal, where residents can view current conditions, historical trends, and health guidance for their specific location.

What 84 Sensors Reveal

The density of our network reveals spatial patterns that 6 regulatory monitors cannot detect. Key findings from 2024 data include the following.

PM2.5 concentrations vary by up to 340 percent within neighborhoods that share a single DEQ monitoring zone. The highest concentrations cluster along major freight corridors, near industrial facilities, and in areas with high traffic density and limited tree canopy. The lowest concentrations are found in well-canopied residential areas away from major roads.

Our sensors identified 14 locations where annual average PM2.5 exceeds the EPA's recently tightened standard of 9.0 micrograms per cubic meter, despite the nearest DEQ regulatory monitor showing compliance. These exceedance locations are predominantly in industrial and mixed-use zones in North Portland, the Brooklyn neighborhood, and along Powell Boulevard.

During wildfire smoke events in August and September 2024, PM2.5 concentrations across our sensor network ranged from 38 to 124 micrograms per cubic meter during a single 24-hour period. The highest readings were in East Portland neighborhoods where housing stock is older and air infiltration rates are higher, meaning that indoor air quality closely tracks outdoor conditions.

When you measure pollution at 6 points, you get a regional average. When you measure it at 84 points, you get a map of who is breathing what. That map looks very different depending on where you live.

Health Implications

We collaborated with the Multnomah County Health Department to overlay our air quality data with health outcome data at the census tract level. The correlation between our sensor-derived annual PM2.5 exposure estimates and asthma emergency department visit rates is striking, with an r-squared of 0.68 at the census tract level.

Census tracts where our sensors show annual PM2.5 above 9.0 micrograms per cubic meter have asthma ED visit rates 2.8 times higher than tracts below 6.0 micrograms. Cardiovascular disease hospitalization rates show a similar, though less pronounced, gradient with a 1.9-fold difference between the highest and lowest PM2.5 quartiles.

These health disparities map directly onto the demographic patterns identified in our Environmental Justice Index. Communities of color and low-income communities are disproportionately located in areas with higher PM2.5 exposure.

Expanding the Network

With funding from the EPA's Community-Scale Air Toxics Ambient Monitoring grant program and the Collins Foundation, TerraFuture is expanding the network to 120 sensors by the end of 2025. The expansion will add monitors in three underserved areas: East Portland beyond 122nd Avenue, the Columbia Corridor industrial zone, and the Cully neighborhood.

We are also adding 12 sensors that measure nitrogen dioxide and ozone, pollutants not captured by our current PM2.5 sensors, at locations near major roadways where traffic-related air pollution is the primary concern. These sensors will use Clarity Node-S technology and undergo the same colocation calibration protocol as our PM2.5 monitors.

The ultimate goal is to produce air quality data at a resolution that matches the scale at which people actually experience pollution, which is their block, their school, their park. That requires density that regulatory networks were never designed to provide. Community science fills the gap.

MC
About the Author
Marcus Chen
Director of Research

Marcus Chen leads TerraFuture's research division, specializing in geospatial analysis and urban ecology. With a PhD in Environmental Science from the University of Washington, he has published over 30 peer-reviewed papers on urban environmental systems.