Skip to main content

Cheaper air quality sensors arrived just in time for the climate catastrophe

It’s gotten a lot easier for anyone to track wildfire smoke and air pollution 

Share this story

A man crosses a street in downtown Portland, Oregon where air quality due to smoke from wildfires was measured to be amongst the worst in the world, September 14, 2020.
Photo by ROBYN BECK/AFP via Getty Images

As smoke from wildfires chokes the West Coast, social media has been flooded with crowdsourced maps providing near-real time updates on just how horrendous the air really is. Much of the data are from relatively inexpensive sensors from a company called PurpleAir. They’ve only been available for the past few years, but they’re already changing everything from government maps of air quality to how communities are watching out for each other — and keeping track of the air they breathe. 

PurpleAir sensors, and others like them have become more popular, and more powerful. 

Low-cost air quality sensors that measure particle pollution — including dust, soot, and smoke — have only become available to most non-scientists in the past decade or so, experts tell The Verge. As people push back against polluters in their backyards and cope with fire seasons that have grown increasingly dangerous as a result of climate change, PurpleAir sensors and others like them have become more popular, and more powerful. 

“The power is not in one individual monitoring their house, but in the individual contributing his data, and another individual, and the municipality, and a scientist,” says Núria Castell, a senior scientist who studies new pollution monitoring technologies at the Norwegian Institute for Air Research (NILU). “We put all this data together and then then we have something,” she says. The resulting high-resolution air quality maps can actually lead to better urban planning and cleaner air when it comes to pollution from fires, industry, or other sources. 

In fact, PurpleAir got its start because of a dust problem. Every day, Adrian Dybwad watched dust from a gravel mine sweep downhill and settle just below his home not far from Salt Lake City, Utah. As the mining company made plans to expand, Dybwad wanted to know how much the dust affected air quality. But there weren’t any pollution sensors nearby, and he couldn’t find one on the market that could do the job and didn’t cost thousands of dollars. 

So in 2015, Dybwad, who has a background in surface-mount electronics and computer programming and networking, set out to build his own. The endeavor grew into PurpleAir: a network of more than 9,700 low-cost air quality sensors that feed data into a near-real time global map of air pollution. For the first time this year, the US Environmental Protection Agency and US Forest Service included data from PurpleAir sensors into its AirNow fire and smoke map.

For the first time this year, the US Environmental Protection Agency and US Forest Service included data from PurpleAir sensors into its AirNow fire and smoke map.
For the first time this year, the US Environmental Protection Agency and US Forest Service included data from PurpleAir sensors into its AirNow fire and smoke map.

PurpleAir’s sensors cost less than $280 dollars — not exactly cheap, but still well below what someone would probably pay for a new smartphone. Compared to the heavy-duty equipment that researchers typically use to measure air quality, it’s a steal — those models can cost up to $50,000. 

That huge range in cost reflects differences in how each air quality sensor is built and operated. High-cost, high-accuracy sensors are typically carefully calibrated, bigger, and use up more energy. Some of these sensors collect particles on a filter and then shine beta rays through them to measure mass. Those types of sensors might require a permit to operate, since the beta rays are emitted from some kind of radioactive source, according to Anthony Wexler, who directs the Air Quality Research Center at the University of California, Davis. Other sensors have very small, fine glass fibers with filters on the end that vibrate almost like a tuning fork, Wexler says. The vibration changes with the mass of the particles collected. And then there’s a slower, more old-fashioned way of monitoring particle pollution by weighing particles captured on a filter in a lab. 

PurpleAir’s sensors measure particulates using laser particle counters. “You basically shine a laser through the air and then the particles in the air reflect the light and the detector picks up those reflections,” Dybwad explains. The method is called “light scattering,” and when Dywbad set out to make his first sensor in 2015, the technology to do this was becoming smaller and more affordable.

“A lot better than nothing”

Around that time, citizen scientists in Stuttgart, Germany were also coming up with ways to take air quality monitoring into their own hands. After meeting in the basement of their city library, Ensia reported in 2017, the group developed a user manual for DIY sensors. The project called Luftdaten, German for “air data,” quickly spread across Europe and scattered across countries in other parts of the world. In China, pushback against dense air pollution also propelled the development of sensors that anyone could buy, say Wexler and Castell. 

More expensive sensors are more accurate than anything someone can easily throw in their online shopping cart. But the sensors that professional researchers use are also too expensive to deploy everywhere. “You have a trade off between high accuracy in a limited number of locations and low accuracy in many locations,” Wexler says. Having the low-cost sensors where there otherwise would be none, Wexler says, is “a lot better than nothing.”

For decades, people living next to sources of pollution lacked evidence that could convince regulators and polluters that they were being harmed. Now, that’s changing. “These low cost sensors are great because they empower communities [living with pollution] to be able to assess what’s going on and say to their regulatory bodies, ‘hey, these guys are killing us,” says Wexler. 

“Hey, these guys are killing us”

Ellen Golla bought an air quality sensor from PurpleAir in 2016 for her home in rural Humboldt County, California. She was worried about pollution from residential wood-burning stoves that she didn’t think official government monitors were picking up on from 30 miles away. “I was breathing obviously polluted air, but according to our county officials, the air was clean. I wanted data on the wood smoke here,” Golla, who helped start an organization to raise awareness of wood smoke pollution, wrote in an email to The Verge

A month after Golla installed her sensor, a fire broke out nearby. “The air quality was incredibly bad in our area for several days. But our air quality management district didn’t issue an advisory about the choking pollution in the north of the county. Officially, we had good air quality, because that’s what the monitor 30 miles away was recording,” Golla said. She started posting her own warnings on Facebook and emailing friends to be wary. “I think that was when I first really began to understand the power of citizen science air monitoring,” she said.

Low-cost sensors available commercially are likely not accurate enough yet to be relied on alone to make policy or regulatory decisions, according to Wexler and Castell. They can miss very small particles or confuse water droplets as particles when there’s high humidity. But many are good enough to raise awareness on air pollution and get the ball rolling to take action. When air quality is really bad — like it’s been across the West Coast of the US over the past couple weeks as a result of wildfires — readings don’t need to be perfect to let people know they should take shelter. 

“I’m very positive that technologically we will advance.”

Castell is also confident that low-cost sensors will become more accurate — whether because of changes to the equipment itself or algorithms that can correct the sensors’ biases. The new AirNow map, for example, applies correction equations to data from PurpleAir sensors that it uses to populate its map. The EPA says it turned to PurpleAir because of how large its network had grown, and the agency tested the lower-cost sensors against regulatory monitors to evaluate their performance. The combined map is a pilot, and its makers are still figuring out how to make the tool better.

“The [low-cost] sensor technologies can really play a big, big role if we take care of the limitations,” Castell says. “I’m very positive that technologically we will advance.”

For Dybwad, who still lives in the same home where he made his first sensor, it’s a “strange feeling” to see hundreds more people purchase his sensors every day as wildfires continue to rage and smoke inundates huge swathes of the West. He checks a Facebook group for PurpleAir sensor owners at least once or twice a day, and still answers users’ questions. “It’s a tragedy, what’s happening,” Dybwad says. 

Update 6:45pm ET: This story has been updated to clarify that Ellen Golla helped found an organization that raises awareness of wood smoke pollution.