While the smoke from forest fires is choking the west coast, social media has been flooded with crowdsourced maps that provide near-real-time information about how awful the air really is. Much of the data comes from relatively inexpensive sensors from a company called PurpleAir. They’ve only been available for a few years, but they’re already changing everything from government maps of air quality to the way communities look after each other – and track the air they breathe.
According to experts, low-cost air quality sensors that measure particulate pollution ̵
“Power is not in one person overseeing their home, but in the person contributing their data and another person, the community and a scientist,” says Núria Castell, a senior scientist who works at the Norwegian Institute for New Technologies Investigated for Air Pollution Monitoring (NILU). “We put all this data together and then we have something,” she says. The resulting high resolution air quality maps can actually lead to better city planning and cleaner air when it comes to pollution from fire, industry, or other sources.
In fact, PurpleAir started because of a dust problem. Every day Adrian Dybwad watched dust sweep downhill from a gravel mine and settle directly under his home not far from Salt Lake City, Utah. When the mining company made plans to expand, Dybwad wanted to know how much the dust was affecting air quality. But there weren’t any pollution sensors nearby, and he couldn’t find one in the market that could do the job and didn’t cost thousands of dollars.
In 2015, Dybwad, who has a background in surface mount electronics as well as computer programming and networking, set out to build his own. The company evolved into PurpleAir: a network of more than 9,700 low-cost air quality sensors that feed data into a near real-time world map of air pollution. For the first time this year, the US Environmental Protection Agency and US Forest Service have included data from PurpleAir sensors on their AirNow fire and smoke map.
PurpleAir’s sensors cost less than $ 280 – not exactly cheap, but still well below what someone would likely pay for a new smartphone. This is a bargain compared to the high-performance devices researchers typically use to measure air quality – these models can cost up to $ 50,000.
This enormous cost range reflects differences in the design and operation of the individual air quality sensors. High-priced, high-accuracy sensors are usually carefully calibrated, are larger, and use more energy. Some of these sensors collect particles on a filter and then pass beta rays through them to measure mass. Anthony Wexler, who heads the Air Quality Research Center at the University of California at Davis, may require an operating permit to operate these types of sensors because the beta rays come from a radioactive source. Other sensors have very small, fine glass fibers with filters at the end that vibrate almost like a tuning fork, says Wexler. The vibration changes with the mass of the collected particles. And then there’s a slower, more old-fashioned way of monitoring particulate pollution by weighing particles that have been captured on a filter in a laboratory.
PurpleAir’s sensors measure particles using laser particle counters. “Basically, you beam a laser through the air and then the particles in the air reflect the light and the detector picks up these reflections,” explains Dybwad. The method is called “light scattering,” and when Dywbad made its first sensor in 2015, the technology for it became smaller and more affordable.
At this time, citizen scientists in Stuttgart also developed ways of monitoring air quality yourself. After meeting in the basement of their city library, Ensia reported in 2017, the group developed a user manual for DIY sensors. The air data project, which stands for “air data”, quickly spread across Europe and across countries in other parts of the world. In China, the backlash against dense air pollution has also spurred the development of sensors that anyone can buy, say Wexler and Castell.
More expensive sensors are more accurate than anything anyone can simply throw into their online shopping cart. However, the sensors that professional researchers use are also too expensive to be used everywhere. “You have a tradeoff between high accuracy in a limited number of locations and low accuracy in many locations,” says Wexler. According to Wexler, having inexpensive sensors where there wouldn’t be any is “much better than nothing”.
For decades, people living next to sources of pollution have lacked evidence to convince regulators and polluters that they are being harmed. It changes. “These inexpensive sensors are great because they empower communities [living with pollution] To be able to judge what’s going on and to tell their regulators, “Hey, these people are killing us,” says Wexler.
Ellen Golla bought a PurpleAir air quality sensor for her home in rural Humboldt County, California in 2016. She was concerned about pollution from wood-burning stoves in residential areas that she did not believe official government monitors were picking up from 30 miles away. “I obviously inhaled polluted air, but our district officials say the air was clean. I wanted data on wood smoke here, ”wrote Golla, who founded an organization to raise awareness of wood smoke pollution, in an email to The edge.
A month after Golla installed her sensor, a fire broke out nearby. “The air quality in our region was incredibly bad for several days. However, our air quality management district gave no indication of asphyxiation pollution in the north of the county. Officially, the air quality was good, because that’s what the monitor 30 miles away recorded, ”said Golla. She started posting her own warnings on Facebook and emailing friends to be careful. “I think that was when I really began to understand the power of civil science aerial surveillance,” she said.
According to Wexler and Castell, commercially available, low-cost sensors are likely not yet accurate enough to allow political or regulatory decisions to be made on their own. You can miss very small particles or mistake water droplets for particles in high humidity. But many are good enough to raise awareness of air pollution and get the ball rolling for action. When air quality is really bad – like it has been on the west coast of the US for the past few weeks due to forest fires – the readings don’t have to be perfect to let people know they should take shelter.
Castell is also confident that low-cost sensors will become more accurate – whether because of changes to the equipment itself or because of algorithms that can be used to correct the distortions in the sensors. For example, the new AirNow card applies corrective equations to data from PurpleAir sensors that are used to populate the card. The EPA said they reached out to PurpleAir about the size of their network, and the agency tested the lower-cost sensors against government monitors to evaluate their performance. The combined card is a pilot project and manufacturers are still considering how to improve the tool.
“The [low-cost] Sensor technologies can really play a big part if we take care of the constraints, ”says Castell. “I am very confident that we will move forward technologically.”
For Dybwad, who still lives in the house where he made his first sensor, it is “a strange feeling” to see hundreds more people buy his sensors every day while wildfires continue to rage and smoke that cover huge chunks of Flood west. He checks a Facebook group for PurpleAir sensor owners at least once or twice a day and continues to answer user questions. “It’s a tragedy what happens,” says Dybwad.
Update 6.45pm ET: This story has been updated to reflect that Ellen Golla helped start a business organization this raises awareness of wood smoke pollution.