After the Smoke Clears: Wildland–Urban Interface Fires and Residues in Nearby Homes

He and other investigators are zeroing in on residences, although other buildings in the WUI—workplaces, churches, schools—may also become contaminated with pollutants. People often have an emotional attachment to their homes, and many have little choice but to return to them afterward; some spend more time at home, as well. “Understanding how building materials interact with chemistry, I realized, really matters,” says Delphine Farmer, another Colorado-based researcher, who also had friends affected by the Marshall Fire. “You want to know what to do about your house when it’s contaminated.”

Through University of Colorado funding and a National Science Foundation Rapid Response Research grant, de Gouw and his colleagues were able to start collecting indoor dust and air samples from residents’ homes just weeks after the Marshall Fire struck. The fact that nearby Boulder has a respected research institution and a population that includes many who work in the sciences led de Gouw to muse on the serendipitous nature of this particular study. “It impacted a community of environmental researchers that is among the best in the world,” he says. “There were all these tools and resources [we needed] to do this research.”

More than $\mathrm{2,000}\mathrm{km}$ (1,300 miles) to the north, Arthur Chan, an atmospheric chemist at the University of Toronto, and his colleagues also conducted research in the aftermath of a WUI fire: the 2016 Fort McMurray wildfire in Alberta, Canada, which destroyed about 2,400 homes and buildings.⁵ Chan says the impetus for their study, which centered around gathering and analyzing dust samples from homes that survived the blaze, came after Alberta Health Services tested the soil in burned neighborhoods and found high levels of arsenic.⁶

The resulting news stories^7,8 about those findings caused residents to wonder if it was safe to go back into their homes. So Chan and his colleagues proposed taking dust samples to analyze for metals and polycyclic aromatic hydrocarbons (PAHs), a class of chemicals produced when biomass burns that can be toxic even at low levels.⁹

But the team knew that by the time they secured funding, residents would have had their houses cleaned, making it more likely that elevated levels of toxicants would have decreased. So instead, they decided to focus on what substances might persist, despite cleaning, in certain areas of affected homes.

For their first study,¹⁰ the team gathered dust samples from bedrooms 14 months after the fire and analyzed them for lingering toxic metals and metalloids, such as arsenic, along with PAHs. A subsequent study, conducted 2 years after the fire, also included samples from the same homes’ entryways and basements, as well as from additional exposed homes.⁵ “We were really concerned about whether these toxicants are going to stay in the home for years and go undetected,” Chan says, “and whether that is going to pose long-term hazards to the people who live there.”

Despite the different timelines, both Chan’s¹⁰ and de Gouw’s³ analyses showed that PAHs and toxic metals and metalloids in the dust they measured were not much higher than background levels in homes nearby but outside the burn area.^11,12

However, de Gouw says they are now finding (in the unpublished work) that the particular PAHs in household dust of nearby homes differ from both PAHs produced by wildfires that burned few, if any, structures and PAHs in homes outside of smoke plumes. Moreover, he says, after cleaning crews worked in one house, the researchers found that fine particulate matter (${\mathrm{PM}}_{2.5}$) from soot and ash became resuspended at levels far higher than, for example, the World Health Organization (WHO) acute exposure guideline for ${\mathrm{PM}}_{2.5}$ of $15{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$.¹³

de Gouw’s team is also finding hundreds of different volatile organic compounds (VOCs)—including the carcinogen benzene—circulating inside the affected houses at concentrations somewhat higher than background levels in unaffected homes. Six weeks passed before those VOCs dropped down to baseline levels. During that time, some homeowners had run simple air filtration devices made of carbon-based furnace filters (which are not designed to remove VOCs) combined with box fans.^14,15 When the devices were turned off, de Gouw notes, VOC levels returned to where they had been prior to running the fans. “Scientifically speaking, it is incredibly interesting,” he says, that levels could rebound, because it suggests there are reservoirs of VOCs in the home. “We don’t know where they are,” he adds. “I wish we did, because then we could clean or remove them.”

That VOC rebound led de Gouw to his current work burning small amounts of synthetic materials in the lab, then measuring the pollutants released in the smoke. de Gouw is conducting his research in the laboratories of Shantanu Jathar and Christian L’Orange, mechanical engineering professors at Colorado State University (CSU). In 2022, Jathar, L’Orange, and colleagues began studying the gases and particles emitted by combustion of common structural materials to develop an inventory of substances released initially, as well as others that are created when those substances intermingle during burning. The authors hope that once their work¹⁶ is published, it will help investigators understand the unique components of smoke that emanates from WUI structural fires.

Chan, too, has turned to burning residential building materials such as plastic—used in sewage pipes, electrical conduits, insulation, flooring, and myriad other items¹⁷—to get a better sense of how their chemical constituents might show up in WUI fire smoke. His pivot from studying potentially toxic household dust after a wildfire to what is in the smoke during the active burn was triggered in part, he says, by the unusually large number of people affected by the Canadian 2023 wildfires,¹⁸ which devastated a total land area larger than Greece.¹⁹

Those Canadian wildfires, which also affected air quality in a large swath of the United States,²⁰ led several journalists to reach out to Joseph Allen, director of Harvard University’s Healthy Buildings Program.^21,22 Allen used these opportunities to stress that during WUI fires, the public should be just as concerned about indoor air quality as outdoor air quality, because smoke penetrates indoors. He recommended using inexpensive indoor air filters (such as those shown above) to clean the air.¹⁵

Then, in August 2023, a wildfire devastated Lahaina, Hawai'i, killing more than 100 people²³ and damaging or destroying more than 2,200 structures, of which the vast majority were residences.²⁴ The Lahaina fire was different from many other WUI fires, Allen says, because such dense residential areas abutted the burn zone. Watching from afar, he saw a situation where his knowledge and expertise might be of use to residents whose homes somehow escaped the blaze.

Allen and colleague Parham Azimi teamed up with researchers at the University of Hawai’i at Mānoa and community members. Their goal, says Allen, is to study “off-site resuspension,” that is, whether particulate matter and VOCs that have contaminated the soil adjacent to burned areas become resuspended when heavy equipment moves through neighborhoods to dig out foundations of destroyed homes and perform other clean-up tasks. They will also analyze whether such pollutants then infiltrate homes that were not destroyed. “We’re interested in making sure people who live in homes that are adjacent to these areas have the tools and knowledge to keep themselves and their families safe,” Allen says.

From January to March 2024 the group took air measurements and collected dust samples from volunteer households in the community; each participant also received a special air filter that captures both VOCs and particles. The investigators are studying whether those devices can sufficiently lower any pollutants that infiltrate and, as in de Gouw’s work, whether there is any rebounding effect when they are turned off. Allen says the goal is to see levels of ${\mathrm{PM}}_{2.5}$ drop to the WHO annual guideline (that is, $5{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$²⁵) or the US Environmental Protection Agency (EPA) primary annual standard of $9{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$.²⁶ “The first questions we’re answering are, ‘What do we find? And are we finding it at levels that are harmful?’” he says. “And then, can we mitigate that on the property? On adjacent properties? On properties that are miles away?”

In March 2022, a few months after de Gouw gathered his samples from the Marshall Fire, Farmer, a professor of atmospheric chemistry at nearby CSU in Fort Collins, conducted a controlled experiment of indoor air quality after a simulated WUI fire. For that study, called the Chemical Assessment of Surfaces and Air,²⁷ Farmer and her colleagues used a test house in Gaithersburg, Maryland, built by the National Institute of Standards and Technology especially for such experiments.

The team created a “typical” indoor soup of airborne chemicals by carrying out household chores such as cooking and cleaning. Then they injected common outdoor pollutants, such as ozone, and compounds to represent off-the-shelf pesticides, such as insect spray, that people might keep in their homes. After taking air measurements to use as a baseline, they began injecting small amounts of a wildfire smoke proxy (produced by burning a mix of needles from common fir and cedar species). In the hours that followed, the researchers observed that levels of VOCs from the injected smoke slowly decreased, but never to baseline levels.²⁸

What the findings tell us, Farmer explains, is that when you have a lot of VOCs indoors, some of those molecules form a weak bond with the molecules on walls, ceilings, furnishings, and other surfaces and stick to them. “And so your entire house [may be] covered in a layer of organics,” she says, explaining that they gradually revolatilize.

The good news is that Farmer and her team found simple cleaning activities, such as wiping off surfaces and vacuuming, helped reduce the VOC levels. Because some surfaces (such as ceilings) could not be easily cleaned, not all VOCs could be readily removed.²⁸ “What we don’t know,” Farmer says, “is how important that is. Is there a health effect?” Research is underway by Underwriters Laboratory Chemical Insights Research Institute to characterize this complex pollution, measure human exposure levels, and determine human toxicity.²⁹

Elliott Gall, an associate professor who directs the Healthy Buildings Research Laboratory at Portland State University in Oregon, took a different approach. He exposed different types of materials found in homes—glass (in the form of Petri dishes), fabric (cut from an organic cotton bedsheet), and mechanical air filter media (like that used in furnace filters)—to 9 hours of a wildfire smoke proxy. Then the materials aged, sitting indoors for 4 months as the scientists regularly measured levels of a class of semi-volatile PAHs associated with wildfire smoke. Gall says one reason the study³⁰ focused on this class of PAHs and not VOCs is because their lower volatility meant they were more likely to stick to indoor surfaces.

He and his coauthors found that elevated levels of these PAHs lasted for roughly 40 days on the samples but were lowered dramatically by simple cleaning steps, like using glass cleaner and laundering the fabric.³⁰ Again, the concern was with surfaces more difficult to clean because these PAHs may also stick to—or even potentially bond with—other household surfaces,³¹ he says, such as carpet or drywall. People may then be exposed through inhalation when PAHs revolatilize, or through skin contact, as may happen when an infant crawls on a carpet. “What we were able to show is that even with a limited exposure to a wildfire smoke proxy, for a very limited number of materials, and with a hypothetical exposure scenario, you still get to health-relevant exposure levels,”³² Gall says.

In yet another approach to studying emissions from WUI fires, US EPA researchers Stephen LeDuc, Amara Holder, and their colleagues analyzed data from two agency air quality monitoring systems: ${\mathrm{PM}}_{2.5}$ Chemical Speciation Network (CSN)³³ monitors, located in major urban areas, and Interagency Monitoring of Protected Visual Environments (IMPROVE)³⁴ monitors, which operate in national parks and other wilderness areas.

Both networks gather data on specific constituents found in particulate matter, such as sulfate, nitrate, and organic and elemental carbon; the CSN network also gathers data on trace metals, such as lead, cadmium, copper, nickel, and iron.³⁵ The researchers chose to conduct their study in California, which has the highest number of speciation monitors³³ and had the most wildfires in the United States during the past 5 years (except in 2022 when it ranked second, behind Texas).³⁶

Holder, a mechanical engineer in the US EPA Office of Research and Development, says that up until about 15 years ago, most large wildfires burned only a small number of structures, which US EPA scientists did not think would appreciably alter the smoke composition of wildfires on a national scale. But then things began to change. “We were seeing that these fires were getting bigger and [involving more homes],” Holder says. “And we began to think we really should try to understand what the emissions are from those types of fires.”

To do that, the researchers used the National Oceanic and Atmospheric Administration Hazard Mapping System for fire and smoke products to identify wildfire smoke plumes in California between 2006 and 2018. Then, they compared data from days when a smoke plume covered an area within range of a chemical speciation monitor with data from days when there were no smoke plumes in that area. They also conducted case studies of WUI fires during that time, to see if smoke plumes from those fires had different chemical signatures. The team chose fires in a range of sizes, from those that burned 100 or fewer structures up to the 2018 Camp Fire, which burned more than 18,000 structures.³⁷

They found an association between the WUI fires that burned the most homes, cars, and infrastructure and elevated levels of copper and lead; at one monitoring station on Point Reyes, California, lead levels were, on average, more than 40 times higher on days impacted by smoke from the Camp Fire than nonsmoke days.³⁸ Such findings intuitively make sense, LeDuc and Holder say, since thousands of vehicles with lead batteries burned during the Camp Fire along with thousands of homes containing lead solder on copper fixtures and pipes. Holder and her colleagues estimated³⁹ that some particularly destructive fires may emit more of these hazardous compounds than all other sources in the airshed. “There are just lots of different [compounds] in a modern world that, if liberated by fire, could enter the smoke plume,” Holder says.

In the United States, federal agencies^40,41 are currently funding laboratory studies to better understand the chemical compounds that come from burning material. Some research, like that of de Gouw and Gall,³⁰ involves combusting small amounts of building materials. Other studies will focus on materials combustion at larger scale. For instance, Holder says her agency has teamed up with researchers from the US Forest Service’s Forest Product Laboratory at Oregon State University, along with engineering consultancy Arup, to burn the contents of a fully furnished small room. “It’s really important to get the different scales,” says Holder, who recently served on the National Academies of Sciences, Engineering, and Medicine committee that produced a consensus study report titled The Chemistry of Fires at the Wildland–Urban Interface.² “The combustion chemistry is strongly impacted by the elements that are present,” she explains.

The issue of scale is something Jathar and his colleagues are also interested in. In June 2024 at the National Fire Research Laboratory, also in Gaithersburg, Maryland, they burned fuel proxies that mimicked, on a larger scale and at greater density, what de Gouw and others had burned the previous summer individually and in smaller quantities: carpet, wood, manufactured wood, and plastic. These new data, which are still being analyzed, could show whether size and density influence emission types or complexity in something closer to a real-world fire. Jathar notes that this work is still far from approximating a whole house fire.

“The small fires that we started in the lab last summer were like a single piece of wood being thrown into a campfire,” he said. “The fires that we’re doing now are like a big, upholstered chair… still orders of magnitude smaller than what you would get in a raging house fire, but we’re going up in scale and complexity.”

It is too early to say what that work will reveal, and Holder says many more studies are needed to get a handle on how the increasing number of large WUI fires affects air quality and human health. de Gouw, who is still analyzing his data from the 2023 project, has noticed that some of the materials gave off toxicants he did not expect to see, such as benzene. And they found chlorine compounds while burning carpet and electrical wiring. Eventually, he says, he will compare his data with the air samples from the Marshall Fire to see if there is overlap. “I don’t have any conclusions to share yet,” he says. “It’s early days, and some of [the comparisons] will be hard to tease out of the data.” He and the other researchers interviewed for this story say their ultimate goal is to offer homeowners—whether in the fire zone or hundreds of miles away—actionable information to protect themselves during and after future WUI fires.

Many of the investigators working in this field live in regions affected by wildfire smoke, so they feel a personal connection to solving this problem. “I have friends who were affected by the Marshall Fire,” says Farmer. “They were asking, ‘Should we repair our carpets? Do we rip out our drywall?’ And I remember thinking naïvely, ‘Oh, there must be literature on this.’ But there wasn’t. That’s one of the reasons we wanted to start doing [these studies].”

For de Gouw, the experience has been one of the most profound of his career. When the Marshall Fire first began, he had the fleeting notion that it was a research opportunity, then quickly squelched it. “This is a tragedy. People lost their homes,” he thought at the time. “They don’t need researchers running around.” Then the questions from homeowners started pouring in, and de Gouw found himself standing in people’s living rooms, waiting for his equipment to finish taking measurements as he listened to their harrowing stories.

“For these people who had been worried for weeks, it was the first time they had someone there to help with their indoor air quality concerns,” he says. “It was very important that somebody listened, took their concerns seriously, and took data.”

He and his colleagues never lost sight of the fact they were dealing with a tragedy. When the team received a 2022 Colorado Governor’s Award for High-Impact Research⁴² and were asked to pose for a picture, they decided it was inappropriate to smile for the camera. As de Gouw puts it, “How do you celebrate an award for work done on a tragedy?” One way is to view the Marshall Fire as an impetus to head to back to the laboratory and keep working, which is exactly what de Gouw and many of his colleagues in atmospheric chemistry are doing.