PurpleAir online mapping of the East Bay at the end of January shows real-time air quality data from sensors around the region. Image captured on 1/26/19 from PurpleAir.com.

“If we can’t see or smell fine particulate matter in our air, it’s hard to be concerned about it, but it can still harm our health,” warned Jim Leach in an online presentation targeted at his neighbors in Lafayette. In his YouTube video posted last August, he urged them to install small monitoring devices outside their homes, schools, and offices to detect worrisome levels of fine particulate matter. Leach is part of a growing movement, sometimes referred to as “citizen science,” in which ordinary individuals use smaller, more affordable devices to monitor and assess their environment.

In his case, Leach relied on a device made by a grassroots group calling themselves PurpleAir, a reference to the “very unhealthy” category on the Environmental Protection Agency’s Air Quality Index scale. He first purchased the device to explore local movement of wildfire smoke, but once he started exploring the data mapped by PurpleAir’s network of devices, he became aware of the many ways in which air quality can vary within a community. Lafayette may be a leafy suburb, but it is bisected by a busy freeway and its quiet neighborhoods lie in valleys that can trap smoke. PurpleAir’s online mapping showed higher levels of pollution in areas near the freeway and in certain neighborhoods, levels that might concern a resident planning to jog or school athletes practicing outdoors.

Leach is participating in the simplest form of hyper-local air quality monitoring. Propelled by leaps in technology and data communication, there is now a plethora of monitoring devices available for a few hundred dollars, rather than tens of thousands, opening up both new ways to assess our environment and new challenges involving what to do with that data.

The California Air Resources Board webpages on community monitoring reveal that many factors must be considered, such as the number of pollutants a device can measure, how it communicates its data, and whether that data is compatible with other monitoring systems. The issue of compatibility is important in order for monitoring data to be compared with or integrated into regulatory programs.

Fern Uennatornwaranggoon, Bay Area Air Quality Policy Manager for the Environmental Defense Fund (EDF), said her organization recognized the need to better capture air pollution at a very localized level, and saw an opportunity presented by the proliferation of new environmental sensors making it possible to collect data at previously unachievable scales. But methodologies to manage and use the new data didn’t exist.

After using different instruments and techniques in a range of projects, EDF is developing and publishing data standards guidelines. “EDF started this program out of necessity,” she explained. “It was a completely new field; no one else was doing it.” They have addressed how to format data (i.e., adding a time stamp), but guidelines on how to calibrate instruments to ensure data reliability and consistency, how to maintain quality assurance and quality control, and how to adjust for changes in weather and over time are still in the pipeline, Uennatornwaranggoon said, adding, “We are catalyzing a space, trying to create infrastructure that can be used by others. Our goal is to make the data findable, accessible, and usable by people, but we’re still a ways off.”

How people will use the data is also still evolving. Like Leach, people can buy a monitor that measures one pollutant, or several at a time. They can use the data to guide their daily activities, or as members of a group that is pressuring regulators for greater pollution controls.

Or they may live in a community like West Oakland or Richmond, where there are a number of sources of pollution, but most residents can’t afford to purchase their own devices. These two communities are improving their monitoring by participating in the Bay Area Air Quality Management District’s Community Health Protection Program, created by 2017’s Assembly Bill 617 (C. Garcia).

It might seem tempting to simply blanket a community with small individual devices and utilize those reporting networks. However, laboratories for the South Coast Air Quality Management District, the largest air quality regulator in the state, report error rates as high as 30 to 50 percent for some sensors. And there’s another problem. Eric Stevenson, meteorology and measurement director at the Bay Area’s Air District, commented, “You can’t just put PurpleAir all over. That may be a starting point, but in order to fully characterize a small area, you’d need hundreds of sensors, which is expensive even though each sensor is relatively inexpensive.”

Where to place them matters, too. EDF’s Uennatornwaranggoon said these efforts face a complex question: “How do you determine the most representative location or area in a community?”

One option is mobile monitoring. Google, EDF, the University of Texas at Austin, and environmental tech company Aclima conducted an innovative one-year air quality streetmapping study in West Oakland ending in May 2016. New compact sensors measuring black carbon and nitrogen oxides were installed in Google Street View cars, which repeatedly drove a set pattern of routes throughout the community. The sensors collected extremely large amounts of data from the constant sampling, which were matched with Google mapping to pinpoint hyper-local variations in air quality within the community. By comparing results to existing Air District monitors in the same area, as well as to other information on pollution sources, researchers confirmed that the sensors were showing places, and points in time, when pollution was highest in certain locations.

Melissa Lunden, chief scientist at the San Francisco-based Aclima, is excited about the potential of using streetmapping for neighborhood monitoring. “It’s a way for a community to make sure that there are answers for the questions that are being asked, without needing to hire their own computing experts,” she observed, explaining that the vast amounts of data acquired by the sensors requires “meta data centers” to process the information. “It allows a community member to take the information and create their own story that they can then use for action.” In West Oakland, streetmapping verified a local pollution source, a metal recycler. “Community members are talking about moving the facility so that it’s not in a residential neighborhood, to where it can expand and use new control technology,” Lunden reported.

Aclima will continue to work with West Oakland representatives on their monitoring plan and is open to participating in Richmond. “We are talking about how to design a sampling program to give reliable standard results, and we would suggest options to the community,” Lunden said. “Our program is easily deployed to map an area of any size. It’s flexible and mutable.”

Stevenson noted that unlike West Oakland, Richmond is just getting started. Following an Air Quality summit scheduled for February 16, which will educate the community about the issues, a steering committee will guide the design of a monitoring plan. “The Air District will start with asking ‘What are your biggest concerns? Here’s what we can do to monitor for those concerns’,” said Stevenson. “We have a general idea of the air quality based on our current network,” he explained. “We need to devise a screening method with enough accuracy to ID sources.”

Stevenson emphasized that the focus is on the community rather than the source. “The message is that we want to monitor in your neighborhood, but we want to do it in an efficient and effective way,” he said, adding that “once you have sound data that people can agree on, people respond to it.”

Stevenson said he considers a mobile platform like Aclima’s to be the best solution for screening for local sources of pollution, but he noted that there is still a role for the “citizen scientist,” particularly for specific events like wildfire smoke. While Aclima’s mapping cars tracked the invasion of smoke from November’s fire in Paradise into the region, Lunden reported only a couple of cars on limited routes provided early data, and once smoke was more dispersed, results were not as clear. “PurpleAir monitors are great for seeing changes in air quality moving through the Bay Area. By comparing various locations against each other at a given point in time, they can help with forecasting and tracking movement [of pollutants] through the area,” Stevenson concluded.

Leslie Stewart covers air quality and energy for the Monitor.