2

Outdoor Sources of Indoor Particulate Matter

In the workshop’s first session, three panelists described some of the major outdoor sources of indoor fine particulate matter (PM_2.5). Cesunica Ivey (University of California, Riverside) discussed what some of those sources are and disparities in exposure to those sources across communities. Brent Stephens (Illinois Institute of Technology) reviewed the mechanisms by which outdoor PM_2.5 gets indoors, and Delphine Farmer (Colorado State University) then talked about the chemical transformations that occur when outdoor PM_2.5 interacts with the indoor environment. An open discussion moderated by Kimberly Prather (Scripps Institution of Oceanography at the University of California San Diego) followed the three presentations.

INDOOR PARTICULATE MATTER OF OUTDOOR ORIGIN AND DISPARITIES IN SOURCES AND EXPOSURES ACROSS COMMUNITIES

PM_2.5 arises from multiple sources including transportation, industrial, and agricultural activities; wildfire smoke, and (in coastal locations) marine vessels (Figure 2-1), and models are an important tool for understanding the contributions of those various sources to ambient PM_2.5, said Cesunica Ivey. As a graduate student at the Georgia Institute of Technology, Ivey developed a hybrid source apportionment model that can attribute primary and secondary particulate matter in ambient air to different sources (Ivey et al., 2015). Her model combines a chemical transport model, a nonlinear optimization method, and spatial and temporal interpolations of adjustment

factors to improve the estimates of source impacts across several different sources.

As Ivey explained, methods such as these can be useful for attributing secondary PM_2.5¹ in the model back to ground sources during meteorological phenomena such as persistent cold air pooling (PCAP) events, which occur frequently in the Western United States during winter and specifically in mountain towns. In the Salt Lake Valley, for example, a hybrid apportionment model found that natural gas and gasoline-powered mobile vehicles were the biggest contributors to pollution during a PCAP event, with refineries and smelters also being important contributors (Ivey et al., 2016, 2019). By using these types of results, said Ivey, it is possible to make specific recommendations about how best to reduce PM_2.5 pollution during PCAP events by targeting precursor sources.

In California, the current focus of Ivey’s research, the state’s Air Resources Board’s estimates of the primary sources of PM_2.5 fall into the “miscellaneous processes” category, which includes residential fuel combustion, windblown fugitive dust, and managed burning and disposal. In the South Coast Air Basin,² such miscellaneous sources—specifically cooking and paved road dust—are again among the top sources of PM_2.5. However, said Ivey, while those sources may be the top contributors to PM_2.5 in that region, the main driver for cancer risk associated with air pollution, according to the Multiple Air Toxics Exposure Study IV (Barbosa et al., 2015), is diesel particulate matter, which does not show up as a top source of PM_2.5. Ivey added that the hot spots for cancer risk map to communities that the state identified as being at high risk of exposure to air pollution.³

She explained that California Assembly Bill 617⁴ requires targeted air monitoring and emissions reductions in overburdened communities, such as East Los Angeles, San Bernardino/Muscoy, and the community near Wilmington and Long Beach. The biggest air pollution concerns cited by members of the San Bernardino/Muscoy community are truck idling, truck traffic, a cement plant, auto body shops, and the BNSF Railway railyard. However, the top five sources of PM_2.5 emissions identified in that community are paved road dust, mineral processes,⁵ and off-road equipment—three source categories that reflect the community’s concerns—as well as

___________________

¹ “Secondary” particles are those formed from chemical reactions in the atmosphere.

² The South Coast Air Basin includes the western portions of Riverside and San Bernardino Counties, the southern two-thirds of Los Angeles County, and all of Orange County.

³Southeast Los Angeles Community Emissions Reduction Program Staff Report. 2020. Los Angeles: California Air Resources Board. https://ww2.arb.ca.gov/sites/default/files/2021-04/SELA_CERP_Staff_Report.pdf (accessed August 21, 2021).

⁴ AB 617; C. Garcia, Chapter 136, Statutes of 2017

⁵ Mineral processes include the production of crushed rock, diatomaceous earth processing, asphalt and cement concrete production, and limestone processing.

cooking, residential fuel combustion, and light-duty passenger car emissions (Figure 2-2).

Possible Sources of Bias

Ivey then discussed biases that can confound the results of models that estimate human exposures. Spatial and temporal biases, for example, can confound acute exposure estimates. She noted that models of ozone exposure in the South Coast Air Basin tend to overestimate exposures in the morning and underestimate them in the afternoon. Another major source of bias is the so-called “change of support problem,” which may arise when inferring values of a variable at places different from where those values are observed (Gelfand et al., 2001; Rongerude and Haddad, 2016).

A further issue with modeling is that the exposure assessment methods that are commonly used by researchers do not address the indoor dynamics of exposure (Goldstein et al., 2021). For example, very small and very large ambient particles are removed as they cross the building envelope, changing the size distribution of indoor particulate matter. As a result, indoor particulate matter includes more particles that tend to linger in the air longer. Overall, estimates of the infiltration factor⁶ for PM_2.5 range between 0.3 and 0.8. Ivey noted that global mortality studies often neglect the role of buildings as attenuators and modulators of exposure.

Human mobility can also confound exposure estimates, with people who are more mobile at higher likelihood of having their PM_2.5 exposure misclassified when using a home-based modeling method versus a call-detail record that relies on cellular phone data (Yu et al., 2020). Ivey added that Google Maps’ location history feature can also serve as a tool for characterizing time and activity patterns (Yu et al., 2019), though doing so for exposure purposes requires consent from the tracked individuals.

To study the extent to which human mobility affects disparities in personal exposure to PM_2.5, Ivey and her collaborators conducted a pilot community-based participatory research field campaign in spring 2019. The 18 adult participants in this study—all residents of the Moreno Valley, Redlands, Riverside, San Bernardino, and Yucaipa communities in inland Southern California, with San Bernardino serving as the disadvantaged community (Do et al., 2021)—wore fanny packs for 7 days; the fanny packs contained monitors that measured PM_2.5 levels every 15 seconds and a GPS data logger that recorded their location every 5 seconds. The major finding from this study was that the San Bernardino residents were more at risk from exposures in their homes compared to residents of the Riverside

___________________

⁶ The infiltration factor is the ratio of indoor to outdoor PM, considering outdoor sources only. The measure is discussed in detail later in this chapter.

and Redlands communities, where the poverty rate is about half that of the San Bernardino community. This finding, Ivey said, supports the idea that “when it comes to mitigation, we do need to consider people’s personal behavior.”

Ivey also pointed out that exposures to what she calls high-risk, nonresidential points of interest should be monitored continuously, given that the participants in the pilot study spent some 30 percent of their time away from home. This work, she added, builds a foundation for using such research methods for larger scale exposure assessments, particularly those facilitated by wearable sensors as well as GPS data loggers.

Emerging Challenges

For source mitigation, Ivey noted that “we may have to consider a new paradigm for how we want to address PM_2.5 exposure disparities and therefore mitigation. We will probably have to consider these approaches through a historical lens where we consider the inequities that may confound our approaches, and we can do this by classifying these exposure disparities as either intentional or unintentional.” By unintentional exposure disparity, she meant exposure that is not affected directly by humans, such as the wildfires that are emerging as an important source of air pollution as the climate changes. In this case, mitigation could involve increasing the air tightness of the building envelope and installing portable air cleaners, a subject discussed in the third session of the workshop.

What drives intentional exposure disparities are inequities in land use, said Ivey. In the Inland Empire region of California, for example, several residential communities are next to warehouses and refueling facilities for the Los Angeles natural gas-powered bus fleet. She suggested that mitigating inequities in source-related exposures may require legislation to ensure that disadvantaged communities no longer bear a disproportionate burden of pollution and environmental hazards by way of land use inequity.

To conclude her presentation, Ivey provided four takeaway messages:

1. Air pollution modeling has paved the way for understanding source-receptor relationships and associated health impacts at the population scale.
2. Misclassification and confounders complicate understanding of exposure disparities and implications for health effects and mitigation for underserved communities.
3. Building envelopes are critically important to consider for future exposure-health assessments that rely on ambient source impact data.

4. Mitigation of exposure disparities will require a commitment to sustainable and equitable zoning and development.

OUTDOOR-TO-INDOOR TRANSPORT MECHANISMS AND PARTICLE PENETRATION FOR FINE PARTICULATE MATTER

The big picture for outdoor PM_2.5 monitoring comes from the extensive network of monitors that the US Environmental Protection Agency (EPA) runs, said Brent Stephens (Figure 2-3). These monitoring stations provide accurate data for local sites and to some extent on a regional scale, but there are gaps between the stations. To fill in those gaps, researchers have developed models using personal-level data of the sort that Ivey discussed, he observed, with remote sensing data also adding to the picture (Figure 2-4). To fill local gaps, researchers are using highly local monitoring with low-cost sensors capable of providing data on outdoor PM_2.5 concentrations at the single-house scale (Bi et al., 2020; Chen et al., 2020).

As has been noted, though, Americans spend most of their time indoors, and the few studies that have examined the relationship between outdoor and indoor PM_2.5 levels have found only weak correlations between the two levels (Klepeis et al., 2001; Meng et al., 2005). As Stephens explained, nearly all outdoor air pollution epidemiologic studies fail to account for these facts, leading to what he called exposure misclassification, “where we are not quite sure if a central site or even a highly localized outdoor concentration of particulate matter is a reasonable surrogate for exposure.”

Sampling indoor and outdoor particulate matter simultaneously yields an indoor/outdoor ratio, where the indoor concentration is influenced by both what is emitted or generated indoors and what infiltrates and persists from outdoors. Outdoor particulate matter infiltrates the indoor environment through open windows and other larger openings, as well as through cracks and gaps in the building envelope and mechanical ventilation (Figure 2-5). The infiltration factor, a number between 0 and 1, may be thought of as a measure that depends on eliminating the presence of indoor sources; it results in an estimate that characterizes how well a building buffers against outdoor pollution.

Infiltration Factor

Several underlying mechanisms govern the infiltration factor for any given building, said Stephens. At the highest level, the infiltration factor is a combination of what penetrates from the outdoors, a function known as the penetration factor (Liu and Nazaroff, 2001), the air exchange rate between the inside and outside of the building, and removal of particles by deposition to surfaces, phase changes (from particle to gas, for example), and air

filters or cleaners. The insulation in a home’s walls can act as a filter that removes particles. If a building has particle penetration via mechanical ventilation, as might be the case in a large commercial building, penetration is in part a function of the type of filter the system uses.

The metrics researchers have used to measure particulate matter infiltration factors fall into two main categories, said Stephens. The most common method uses a chemical surrogate of the samples collected on a filter and weighed. Sulfur and sulfate aerosols, for which there are few indoor sources, are a reasonable proxy for the infiltration factor for various classes of particulate matter, including PM_2.5 (Sarnat et al., 2002; Wallace and Williams, 2005), though in some areas outdoor sulfate concentrations have decreased enough that it is becoming less useful as a surrogate.

Time-resolved monitoring with subsequent data processing is another approach for estimating particulate matter infiltration rates. “The idea here is if you measure indoor and outdoor concentrations of a particulate matter of some kind on a time-resolved basis, you will get data showing where periodic episodic indoor peaks shoot up over the background,”

Stephens explained (Kearney et al., 2014; Kunkel et al., 2017; Liu et al., 2019). Identifying the peaks and eliminating their indoor sources provides a means of comparing the indoor/outdoor ratio when there are no emissions indoors. He and a colleague had access to unoccupied apartments where there were with no indoor sources, and they found that very small and very large particles do not penetrate and persist in the indoor environment (Zhao and Stephens, 2017).

Stephens noted that a large survey of studies found that the median infiltration factor for PM_2.5 is approximately 0.5, meaning that the average building filters out about half of the outdoor PM_2.5 (Chen and Zhao, 2011). In comparison, the median infiltration factor for PM₁₀ and ultrafine particulate matter (PM_0.1) is approximately 0.3 (Kearney et al., 2014; Stephens, 2015). The infiltration factor of any one home can vary tremendously, he added. The key drivers of variability in infiltration factors include pollutant characteristics such as size, class, and chemical components of the particulate matter; the source of ventilation air; human behaviors such as

how often the residents open and close windows and doors and their use of portable air cleaners; the magnitude of the air change rate; and the heating, ventilation, and air conditioning (HVAC) system runtime and filter efficiency.

Penetration Factors

A variety of approaches, Stephens continued, can be used to indirectly measure penetration through the building envelope. The most common methods use integrated gravimetric PM_2.5 samples (those collected on some filter medium and then weighed) and a regression analysis across homes, but there are accuracy challenges due to the lack of dynamic data. One study, for example, estimated the penetration factor for homes to be approximately 0.9 (Meng et al., 2005)—with a range of about one order of magnitude.

There is also a robust and growing literature on methods intended to measure penetration factors directly (Peng et al., 2020; Rim et al., 2010; Vette et al., 2001). In one study, researchers elevated particle concentrations in an unoccupied building, let the levels decay to background levels, used a portable HEPA (high-efficiency particulate air) cleaner to scrub the air, and then let the concentration level rebound (Thatcher et al., 2003). This

approach provides a means to determine penetration factor and deposition loss rates simultaneously, but it is time-consuming and requires an empty indoor space. Stephens said he is fortunate that the Illinois Institute of Technology maintains an apartment on campus that he and his team can use for these types of studies. Together, the studies confirm that ultrafine and large particles do not penetrate and persist for long time periods in buildings.

New Directions in Assessing Infiltration and Penetration Factors

An exciting piece of research Stephens has seen recently used a low-cost sensor network to estimate infiltration factors at the scale of the building stock (Bi et al., 2021). This study used nearby outdoor PurpleAir monitors⁷ matched with indoor measurements to develop a distribution of estimated infiltration factors. While this approach needs to be corroborated with data from more established measurement approaches, it is promising. Another new approach uses data from hundreds to thousands of homes outfitted with low-cost sensors to build models based on building stock characteristics, land use, geographical information system parameters, and other factors that can predict infiltrations factors across the building stock (Tang et al., 2018).

The results of these methods, said Stephens, could enable integration of infiltration factors in exposure estimates and ultimately in environmental epidemiology. He and his collaborators, for example, have examined the typical concentration response function for outdoor PM_2.5 and modified it for the underlying exposures to microenvironmental PM_2.5 exposure contributions, both indoor and outdoor (Azimi and Stephens, 2020). They are also examining how to better attribute indoor and outdoor sources across the various microenvironments in which people spend their time. So far, this approach has shown that the single largest source, on average, of PM_2.5 exposure in the population is (by a small margin) outdoor PM_2.5 that has penetrated inside homes; PM_2.5 of indoor origin is the next largest contributor.

A group at Lawrence Berkeley National Laboratory has been studying the effects of mechanical ventilation and filtration on infiltration factors in a single unoccupied residence (Singer et al., 2017a). One finding by this group, Stephens said, was that improving filtration in the house reduces the infiltration factor. However, even with a “good” filter in place, mechanical ventilation that brings in outdoor air will increase the infiltration factor and the concentration of outdoor particles inside the house. “Improving

___________________

⁷ PurpleAir is a manufacturer of air quality monitors. The devices may be linked to a network that displays real-time measurements on a publicly accessible website (https://www.purpleair.com/map).

that supply filter then gets about the lowest infiltration factor you can,” he observed. “This is not rocket science, but it is nice to have some quantitative information on how we can adjust the building stock and what certain retrofits may or may not do to the proportion of outdoor particles that penetrate and persist indoors.” The findings are intuitive, but they do point to approaches that could be used to reduce PM_2.5 exposure indoors.

Over the past few years, investigators have been using health-relevant metrics in assessing infiltration and penetration factors. Stephens mentioned two studies that have examined the oxidative potential⁸ of indoor infiltrated PM_2.5: one in an unoccupied test house (Khurshid et al., 2019) and the other in an unoccupied apartment (Zeng et al., 2021). These studies show that the oxidative potential of particles that infiltrate from outdoors increases indoors, and that this phenomenon is positively correlated with differences in outdoor and indoor temperature and relative humidity.

Stephens concluded his presentation with a list of research needs, which included these:

• Integrate indoor exposure attributions to ambient particulate matter epidemiology investigations to address exposure misclassification and improve the accuracy of health effect estimates.
• Determine the differential toxicity of PM_2.5 of indoor and outdoor origin.
• Address the problem that direct measurements of infiltration factors is expensive at scale. Today, such studies are typically limited to samples of convenience, but there is the potential to leverage advances in low-cost sensors.
• Improve the direct measurement of penetration factors by increasing sample sizes and incorporating PM_2.5 chemical composition, standardizing approaches, and exploring influencing factors.

OUTDOOR PARTICULATE MATTER SOURCES AND THE CHEMICAL TRANSFORMATIONS THAT TAKE PLACE WHEN THEY INTERACT WITH THE INDOOR ENVIRONMENT

Building on the foundations laid by Ivey and Stephens, Delphine Farmer focused on sources of particulate matter in the outdoor environment and what happens to those particles after they infiltrate the indoor environment. Her perspective on particulate matter is to think about it in terms of sources and sinks. She explained that the change in concentration of whatever a

___________________

⁸ Oxidative potential is “the capability of particles to generate reactive oxygen species in a biologically relevant system, [which] may be a useful indicator of the intrinsic toxicity of PM” (Zeng et al., 2021, p. 2).

person breathes as a function of time is equal to the amount produced through direct emissions plus chemical sources arising from reactions in the atmosphere minus the amount lost through deposition and through chemical reactions, and a meteorology term that accounts for the effects of air movement (Emerson et al., 2020; Lee et al., 2013; Li et al., 2020).

Farmer also noted that while deposition is what controls airborne particle lifetime, it is arguably the least understood component in both indoor and outdoor environments. What is known is that for PM_2.5 and smaller particles, the average atmospheric lifetime is about 7 days. “What that means is that what we breathe is a combination of both local sources and sources transported from other regions,” said Farmer (Figure 2-6).

Research demonstrates that particulate matter levels have been decreasing across the United States thanks to the Clean Air Act and regulatory activities that have reduced not only primary emissions but also the atmospheric chemical reactions that produce particulate matter (Jaffe et al., 2020; McClure and Jaffe, 2018). The exception is in the Western United States, where wildfire smoke is increasing the amount of outdoor particulate matter that is primarily organic in composition. This increase in organic aerosol from biomass burning, said Farmer, is influencing both outdoor and indoor air quality.

Several years ago, Farmer and her collaborators put an aerosol mass spectrometer on an airplane, flew it into wildfire plumes, and then traced the smoke plume as it diluted to study the chemistry occurring in the

atmosphere (Garofalo et al., 2019; Lindaas et al., 2021; Palm et al., 2020). The first thing that happens is that the plume dilutes in a manner that affects the equilibrium of the smoke constituents. “The organic aerosol that dominates biomass burning emissions has some semivolatile components, and when you move it from a high-concentration region to a low concentration, equilibrium drives some of those semivolatile components off in dilution-driven evaporation,” Farmer explained. In addition, the trace gases in wildfire smoke get oxidized, and the oxidized volatile organic compounds (VOCs) can have a low enough volatility that they condense and partition to the particle mass to form a secondary organic aerosol.

Furthermore, the volatile components released by dilution-driven evaporation can also oxidize and then condense. In fact, analysis of the trace gas and particle data collected during the project revealed that within a few hours after emission, a third of the particulate matter from smoke was actually secondary organic aerosol formed through this latter process (Palm et al., 2020). “This should tell you that the atmosphere is a dynamic environment, with many chemical reactions, and this is constantly changing the composition, the amount, and the size distribution of the particles that are in the atmosphere,” said Farmer.

The sources of primary particulate matter and precursors to secondary outdoor particulate matter include anthropogenic sources—fossil fuel combustion, for example—and natural sources such as volcanic emissions and sea spray. The biosphere, Farmer added, is a massive source of VOCs that can react in the atmosphere to produce particulate matter. These sources are not evenly distributed across the planet—there is more sea spray aerosol along the coasts than in the Rocky Mountains, to cite an obvious example—so there is spatial heterogeneity in the composition of outdoor particulate matter. One recent study found that particle concentration is driven by crustal material, secondary inorganic aerosols, and biogenic secondary organic aerosol, and that anthropogenic aerosol from sources such as wood burning and vehicle wear drives the oxidative potential in that aerosol (Daellenbach et al., 2020). Oxidative potential, said Farmer, defines how reactive an aerosol is when it interacts with a cell in the human body and the harm it can produce via oxidative stress. “What this study highlights to me is that we need to understand more than just aerosol sources,” she said. “We need to think about the chemistry and use the chemistry to link to health effects.”

Chemistry and Indoor Particulate Matter

As both Ivey and Stephens noted, most people spend the majority of their time indoors, making it important to understand what happens to the chemistry of outdoor particulate matter as it moves indoors. The indoor

environment, said Farmer, has a very high surface area and a large array of volatile compounds compared to the outdoor environment, but very low oxidant loadings. As noted, particles may remain airborne for 7 days outdoors in the absence of influences, but with mechanically ventilated spaces indoors that time shrinks to hours or minutes. In addition, once particles infiltrate the indoor environment, they exist in much lower concentrations, and the more volatile components have lower indoor-to-outdoor ratios (Johnson et al., 2017).

Temperature gradients are an additional factor to consider because particles undergo thermal partitioning as they infiltrate, Farmer observed. When they warm during infiltration they can volatilize, whereas when they move from a warm outside to a cold inside, they can condense. In fact, there are seasonal differences in the indoor-to-outdoor ratios of various VOCs in the particle phase (Avery et al., 2019b). Chemistry comes into play when considering that both human activity indoors and the indoor built environment add a huge amount of volatile compounds indoors. Oxidant levels are low indoors, but they are not zero and the oxidants that are present can react with the organic load to produce an array of oxidized organic and inorganic products that can then condense to form secondary organic aerosol (Avery et al., 2019a).

In one experiment, investigators showed that limonene, a key ingredient of anything that smells like citrus, reacts with household bleach vapors to form new particles (Wang et al., 2019). While this was a laboratory study, Farmer and her colleagues were able to reproduce that finding in a field study in which they showed that the VOCs released during cooking could react with bleach used in household cleaning to produce oxidized organic aerosol (Mattila et al., 2020). She noted that the fraction of particulate matter produced was small, which she believes is driven by the short timescales over which chemistry can occur in indoor environments.

There are many other examples of research demonstrating the important role of indoor chemistry in particle formation. Farmer cited a study in which investigators took particles of different compositions, put them in a room with vinyl flooring—which releases very small amounts of phthalate esters—and identified chemical reactions that could produce aerosols (Eriksson et al., 2020). In addition, this study found that the phthalate esters condensed more efficiently on some particles than others, showing not only that chemistry matters in the indoor environment but that pollutants can move from one place in a house to another. Another study showed that cigarette smoke generated outside can infiltrate indoors, deposit on surfaces, undergo chemical processing on those surfaces, and revolatilize on timescales of months. That so-called “third-hand smoke” can then partition onto different types of particles (Collins et al., 2018; DeCarlo et al., 2018).

In summary, Farmer said that sources of outdoor particulate matter are spatially and temporally variable, and that predicting outdoor, and thus indoor, particulate matter requires understanding emissions, chemistry, and deposition rates. In addition, indoor particulate matter undergoes equilibrium partitioning in terms of dilution and temperature as well as chemical reactions. Given that, she concluded with three open questions and research needs:

1. What is the indoor budget of condensable material (that is, its concentration, reactivity, and volatility)?
2. Which multiphase reactions compete with ventilation and deposition and over what timescales?
3. How do particle composition and phase affect indoor particulate matter chemistry?

Her conclusions on the basis of the studies discussed were that there is broad scientific consensus that outdoor PM_2.5 influences indoor PM_2.5, so reducing outdoor sources improves indoor air quality; and that once indoors, aerosols have many chemical fates—albeit on far shorter timescales than outdoor air—and that the resulting chemical composition of the air may influence human health.

DISCUSSION

Kimberly Prather, the moderator, invited questions from planning committee members and then from the webinar viewers.

Elizabeth Matsui noted that she was struck by the complexity of the outdoor sources, the penetration of the indoor emissions from them, and the chemistry involved in creating particulate matter. She then asked the speakers if there is a framework to identify and prioritize the best targets for mitigation. Farmer replied that while the outdoor atmosphere is complex, many of the underlying fundamental principles are understood, including about the chemistry occurring on regional and global scales. The open questions, she said, are more at the local level that Ivey addressed in her presentation. Though applying those underlying chemical and physical processes to indoor environments will be challenging, it is most likely not an intractable problem.

Stephens said that there is some evidence that factors such as air conditioning usage modifies some of the outdoor air epidemiology associations with PM_2.5, suggesting that filtration and newer building stock can mitigate some of the adverse health effects. He quoted a mantra in building science, “build tight, ventilate right,” but said it requires efficient filters to prevent buildup of particulate matter, and noted that the marketplace is not meeting

the need for effective and energy-efficient ventilation for homes. “None of this is rocket science,” said Stephens. “It is about priorities and incentives.” Addressing the existing building stock is a challenge, he added, in part because it is not well understood.

Rengie Chan asked the panel if there is a need for more research on the efficacy of sealing homes to reduce infiltration and how to translate that into practical mitigation steps. Stephens answered that unpublished data from his group showed that energy retrofits did not have a big effect on infiltration factors in part because of varying meteorological conditions and variability in air sealing. But he added that if he had to make a decision, he would say sealing homes would have a big effect, although “exquisite” data do not exist showing that to be true.

Prather observed that sealing buildings goes against all the advice regarding COVID-19—to “ventilate, ventilate, ventilate”—and she asked the panel to talk about trade-offs of reducing exposure to outdoor air pollutants and increasing indoor exposure to viruses. Stephens responded that good filtration is the key to balancing those two concerns. Farmer noted that the key was to seal a building well and ventilate it properly.

With regard to marginalized and underserved communities, Richard Corsi asked the speakers to comment on the ability of such communities to reduce their exposures when they may not be able to afford to purchase more expensive filters or good, portable HEPA cleaners. Ivey replied that the nation needs to get creative about how to leverage policy and perhaps use programs such as Medicare and Medicaid to cover the interventions that would improve health. “This needs to become a systemic intervention because the causes of disproportionate exposures are systemic,” said Ivey. “They are not individualized causes, and to fix a systemic issue, you need systemic solutions.” Farmer added that the nation needs to think about the different sources of outdoor particulate matter, which are out of an individual’s control, and understand what those sources are and what they contribute to the health burden in disadvantaged communities. In terms of the many opportunities that exist to decrease particulate matter exposure and improve indoor air quality, she noted that while this is a socioeconomic issue, there is a need for the scientific community to do a better job explaining to all communities which approaches are the best for producing the desired results in a cost-effective manner.

Several online participants wanted the panelists to identify what the best metric would be for quantifying acceptable levels of indoor particulate matter. Farmer explained that the challenge of developing standards is having enough information to know what levels people are exposed to, what sources one can actually control, and then what the health effects are of the various constituents of indoor particulate matter. For ultrafine particles, there is a growing body of evidence that they can penetrate deep into the

lungs and produce adverse health effects. In her view, many of the questions about regulations or acceptable levels should be guided by the precautionary principle, the idea of first figuring out that certain levels are safe and being cautious about setting guidelines. “The science is there. It is very clear that it shows that particulate matter has negative consequences to human health,” said Farmer. She added that the science is also clear that there are dangerous compounds in the air that can be produced by chemistry and that some have negative health effects. “I think we can use that information to start to develop guidance,” she said.

Stephens noted that PM_2.5 is often used as a blanket measure because it is harder to get data on composition and size resolution for smaller particles. Epidemiology studies, in particular, struggle in that regard, in part because of the way monitoring and regulatory frameworks have been established. He noted a study from the Center for Nanotechnology and Nanotoxicology at the Harvard T.H. Chan School of Public Health showing that particles generated by photocopiers and printers at a copy center had the potential to harm the lungs of those exposed to them (Pirela et al., 2013). While he characterized the framework for this study as elegant, with the potential to increase understanding of the competing effects of indoor and outdoor sources, he added that it took tremendous resources to conduct that one study at one copy center.

Ivey remarked that there is a need to think about where the greatest disparities exist and to consider who is most negatively affected and how they are most negatively affected, given that society is only as strong as its “weakest” population. One step would be to conduct more surveys in the communities that have the highest exposures and that are surrounded by the sources that potentially lead to cancers, and then establish standards to protect those who are most vulnerable in society. In the same vein, Farmer said that there is a great opportunity to apply the technologies developed to study and understand the association between what happens in the outdoor environment and the indoor environment and do so in the most vulnerable and disadvantaged communities.

An online participant wanted to know what criteria the research community is focused on to quantify the intersection between particle infiltration and health, particularly regarding time-integrated exposure versus instantaneous maximum exposure. Stephens said the literature shows that both are being studied. The most robust associations to build on regarding health outcomes are those on long-term exposures to particulate matter, with few studies on the daily exposure component. He added that there have been only a handful of studies that examine whether infiltration affects emergency room visits, for example. Prather concurred with that assessment, noting that spikes in outdoor particle levels could have important health effects but there have been few studies examining that idea, and

Farmer said the challenge arises because indoor environments may have high ventilation rates and indoor concentrations can change rapidly. This can result in circumstances where there are repeated short-term, high-concentration exposures. “I do not think we really understand how those different timescales of exposure link to health effects,” said Farmer.

The final question to the panel asked if there are any studies on long-term (50- to 80-year) trends in air quality versus human health, wondering if at this point only incremental benefits are being achieved from further, potentially expensive interventions. Stephens said that a 2009 paper described changes in indoor air pollutants since the 1950s (Weschler, 2009), but he was not aware of much more than that. Farmer noted that there are good data on long-term trends for outdoor particulate matter and clear links to changes in hospital visits, mortality rates, and cardiorespiratory disease. The data show that reducing particulate matter exposure by even small amounts has very substantial beneficial health effects, which is why the Clean Air Act is viewed as one of the strongest pieces of legislation in terms of extending life years of people in this country.

This page intentionally left blank.