Abstract

Background

Prescribed burning is an important fuel management tool to prevent severe wildfires. There is a pressing need to increase its application to reduce dry fuels in the western United States, a region that has experienced many damaging wildfires. Public support for this practice is tempered by concern around smoke impacts and escape risks. This study aims to understand how recent experiences with wildfire smoke and perceived risk of smoke events affect public support for prescribed burning.

Methods

Data were from the May 2023 Household Emergency Preparedness Survey, an online panel survey of 1,727 adults in 12 western conterminous states, applying survey weights to reflect the underlying population demographics. In weighted logistic regression models, we evaluated associations between predictor variables (past experiences with smoke, wildfire risk perception) and support for prescribed burns in general or near a respondent鈥檚 neighborhood, adjusting for age, race/ethnicity, gender, education, household income, and wildland urban interface status. Mediation models were used to assess whether perceived risk of smoke exposure mediates the relationship between recent smoke experience and support for prescribed burning.

Results

Approximately two-thirds of the population supported prescribed burning in general, and more than half supported prescribed burning near their neighborhood. 44% reported experiencing a smoke event in the past 3 years, which increased the odds of support for prescribed burning in general (OR鈥�=鈥�2.03, 95%CI 1.51鈥�2.74) and near their neighborhood (OR鈥�=鈥�1.59, 95% CI 1.20鈥�2.09). High perceived risk of future smoke impacts was associated with support for prescribed burns in general (adjusted OR鈥�=鈥�1.66, 95% CI鈥�=鈥�1.15鈥�2.39) and near their residence (adjusted OR鈥�=鈥�1.72, 95%CI鈥�=鈥�1.23鈥�2.39). Although only trending towards significance, perceived future risk mediated 16.9% (p鈥�=鈥�0.066) of the association between recent smoke experience and support for prescribed burning nearby. Among those who experienced recent smoke events, reporting high degrees of overall smoke impacts or outdoor air quality impacts were positively associated with support for prescribed burns.

Conclusions

Recent experience with wildfire smoke and perceived future risk are strongly associated with support for prescribed burns. Educational campaigns can apply these findings to improve public support toward prescribed fire activities and funding to reduce wildfire risks and protect public health.

Background

There is growing recognition that climate change has worsened the impacts of wildfires in the past few decades [1]. With the increases in severe wildfires in the United States (U.S.) [2, 3], the economic burden of wildfires (including the costs of fire suppression and both direct and indirect losses) is now estimated to be between $71.1听billion and $347.8听billion annually [4], including loss of life and impacts to physical and mental health [5, 6]. In the western U.S., prolonged dry periods and strict fire suppression policies throughout much of the 20th century have resulted in a large amount of dry fuel available for wildfires [7,8,9,10]. As urbanization increases and populations expand into the wildland urban interface or intermix (WUI), strategies to mitigate wildfire risks and damages are critical to protecting community health and safety [9]. The health impacts of wildfires include loss of life and disability related to injuries or damage sustained in the immediate vicinity of a wildfire burn event, as well as greater risks of adverse respiratory [5, 11,12,13,14], cardiovascular [12, 15,16,17,18,19], and neurocognitive [20] health conditions, and more [21,22,23,24,25], from acute and chronic exposure to contaminants in air [26, 27], water, and soil emitted from wildfires through smoke and ash. There is increasing awareness of the need to protect the public鈥檚 health from these wide ranging impacts of wildfire smoke while also managing and preventing wildfires [28].

One key tool in combatting the risks of severe wildfires is the controlled application of fire to remove dry fuel, known as prescribed burning [29, 30]. Cultural fires, which have long been used by many Native American tribes, are one category of prescribed fires, but the majority of prescribed fires in the U.S. are those managed by non-tribal fire agencies [31]. Public opinion around prescribed burns in the United States has changed over time, with generally negative opinions expressed from the early 1900s through the late 1980s and beyond, characterizing fire as being destructive to the environment [32]. However, as forest management science developed, the benefits of controlled fire have become more widely accepted and acknowledged, such that modern day land management policies incorporate the use of prescribed burns as a way to mitigate the material damages, exposure and health risks of severe wildfires [33]. Despite this shift toward embracing prescribed burns in fire management policy, there are challenges to expanding its use due to negative public perceptions, limits to funding, crew availability, crew experience, and burning regulations related to weather and season [30, 34]. Prescribed burns do incur some degree of risk, including air quality impacts which, in turn, impact the health of vulnerable populations and the small risk of escape. Building positive public perceptions of prescribed burning will be an important next step to bolster public support for regional prescribed burn policies and sustained funding for this wildland management tool that reduces risks of severe wildfires and thereby mitigates the associated health impacts [31, 35].

Past research into public perceptions of prescribed burning has often focused on specific communities, especially communities that recently experienced major fires, or communities in fire prone areas [33, 36,37,38]. Research on the perceptions of prescribed burning among United States residents suggests that prior wildfire exposure and concern for the environment is associated generally with wildfire risk-reducing actions [39]. Considering research that has focused uniquely on prescribed burns, several studies conducted in the past 20 years suggest that there is strong public support for prescribed burns [34, 40]. These same studies noted that there are public concerns around smoke and control of the fires; however, the sentiment regarding the value of prescribed burning is overall positive with heterogeneity across sociodemographic characteristics [40]. Similar to studies of general preventive practices, more positive sentiment is associated specifically with a recent history of fire, trust in the fire service teams, as well as rurality [41,42,43]. In contrast, people who have experienced negative health effects due to wildfires or are less smoke tolerant due to a health condition reported greater negative sentiments towards prescribed burns [41, 44]. Overall, public acceptance of fuels management practices and community preparedness for wildfires were connected with perceived risk, trust in who is providing information, and other factors [33, 45]. In addition, studies have found that prescribed burns tend to occur in areas that are more rural, have a higher percentage of minority individuals, and with greater degrees of economic disadvantage (e.g. more vacant homes and mobile homes and with lower incomes) [46].

Given the importance of prescribed burns as a fuels management tool to prevent large wildfires, and knowing that public support is critical to help secure additional, year-round governmental funding for prescribed burn activities, understanding factors that shape public sentiment is critical [47]. The current project examined the degree of public support for prescribed burning among residents in the conterminous western United States, which experienced several severe wildfire seasons between 2020 and 2022 [48]. These damaging wildfires led to the Wildfire Crisis Strategy announced in 2022 by the U.S. Department of Agriculture Forest Service, which emphasized the need to increase fuel reduction treatments by four times the current levels in the western U.S [49]. We examined the types of information that may be helpful to these residents and evaluated whether past experience with wildfire smoke or perceptions of future wildfire risk may influence support for prescribed burns.

Methods

Data sources

Data for this study were collected in May 2023 as part of the Household Emergency Preparedness Survey, a survey of a nationally representative sample of adults in the United States. The sampling frame first considered the 4 major census regions (West, South, Midwest, Northeast), with recruitment targets set to provide a demographically representative sample within each region. The survey module pertaining to prescribed burns and wildfire smoke exposures was only provided to respondents living in the EPA Regions 8, 9, and 10, which is the focus of the current study. Sampling quotas for each of these EPA regions were set to achieve statistical estimates with an acceptable margin of error (卤鈥�2.4%), and aimed to collect 400 responses each from EPA Regions 8 (Colorado, Montana, North Dakota, South Dakota, Utah, Wyoming) and 10 (Alaska, Idaho, Oregon, Washington) and 900 responses from EPA Region 9 (Arizona, California, Nevada). Hawaii was not included in the survey design. In addition to geographic stratification, the survey sampling design also invited respondents based on demographic characteristics defined by age, race/ethnicity, household income, and living in a rural/urban zip code, to ensure representation and an acceptable level of statistical precision across demographic groups including traditionally hard-to-reach populations (ages鈥�<鈥�30 or 鈮モ��65, racial/ethnic minorities, rural regions, low-income households). Post-stratification weights were applied to collected data to ensure that the sample demographics reflected the population characteristics of the underlying population, based on the 2021 American Community Survey 1-year estimates for age group, sex, race, Hispanic ethnicity, and rural zip code status, as well as the 2022 Current Population Survey Annual Social and Economic Supplement for household income. The survey instrument was designed in consultation with experts and a survey design team; the survey included modules assessing perspectives, impacts and preparedness pertaining to wildfire smoke, along with other types of emergencies. The Salus Institutional Review Board (IRB) determined that this study qualifies for an exemption from the need for IRB review (Study ID: 23074-01). All participants provided informed consent prior to taking the survey. The panel survey was administered by Dynata.

A total of 1780 survey responses pertaining to prescribed burns and wildfire smoke exposure were collected for the Western region. Forty records were excluded based on data quality assurance criteria (e.g. incorrect zip codes, failed quality check questions, within the bottom 1% of time needed to complete the survey, etc.). Additionally, we excluded 13 responses from Alaska because we did not have wildland urban interface data in that geography. The analytical dataset included n鈥�=鈥�1727 records. Zip codes in the survey dataset were converted to Zip Code Tabulation Areas (ZCTAs) using the 2022 Uniform Data System Mapper crosswalk file [50].

Support for prescribed burns was assessed through two survey questions:

(1) Do you support the use of prescribed burns in general to reduce the risk of wildfires?

(2) Do you support the use of prescribed burns near your neighborhood to reduce the risk of wildfires?

Prescribed burns were described as 鈥渟mall, controlled fires used to reduce the risk of wildfires by burning up dead wood and vegetation鈥�. Responses were provided on a 5-point Likert scale and recoded to binary indicators of support (including 鈥渟trongly support鈥� and 鈥渟upport鈥�) or lack of support (including 鈥渘either support nor oppose鈥�, 鈥渙ppose鈥�, and 鈥渟trongly oppose鈥�) for prescribed burns in general or near their neighborhood.

All respondents were also asked about whether they had experienced wildfire smoke during the previous 3 years (鈥渞ecent wildfire smoke experience鈥�). Among those respondents who indicated they had experienced smoke in this time period, the survey also asked them to rate how much they were impacted by the wildfire smoke on a 10-point Likert scale. The survey assessed overall impacts, as well as specific types of impacts, such as impacts on health, mood and mental well-being, indoor and outdoor air quality, ability to engage in exercise, and transportation. Responses for smoke impacts were categorized as follows: 1鈥�3 (low), 4鈥�7 (moderate), and 8鈥�10 (high).

Wildfire risk perception was evaluated using two questions assessing the risk of being personally impacted and the perception of changes in frequency of wildfire smoke events. To assess perception of future personal risk, the survey asked the respondent to rate on a 10-point Likert scale how much they felt at risk for being personally impacted by wildfire smoke in the next 12 months; we considered responses scoring between 8 and 10 as 鈥渉igh risk鈥�. To assess the perception of changes in frequency, the survey asked whether the respondent thought wildfire smoke events will become more frequent or less frequent in the next 5鈥�10 years, near where they live; responses were provided on a 7-point Likert scale ranging from 鈥渕uch less frequent鈥� to 鈥渕uch more frequent鈥�, with any response indicating an increased frequency considered as 鈥渉igh risk鈥�.

The survey additionally asked about whether the respondent would want additional information about specific aspects of prescribed burns, including information about: (1) how prescribed burns are conducted, (2) what precautions to take when prescribed burns are being conducted, (3) how prescribed burning reduces the risk of wildfires, and (4) when and where prescribed burns are being conducted. Responses were provided on a 10-point Likert scale, and we considered responses between 8 and 10 points as reflecting a strong desire for more information.

The United States Geological Survey WUI map for the conterminous U.S. was used to determine zip code-level WUI percentage [51]. This WUI map was developed based on building point locations (2018 release) and wildland vegetation maps (2016 National Land Cover Dataset). The map displays both the wildland urban intermix and the wildland urban interface, based on Federal Register definitions. The maps were generated using a range of circular buffer distances to define neighborhood size; we selected the 500听m radius buffer distance, which is consistent with the Federal Register definition of the WUI.

WUI data were intersected with the 2020 U.S. census ZCTA Tiger file using ArcGIS Pro (version 3.2.1, ESRI, Redlands, CA) to calculate the percentage area of each ZCTA classified as either wildland urban intermix or interface. We defined ZCTAs to have a high level of WUI if 40% or greater of the ZCTA area was in either the wildland urban intermix or interface. Approximately 25% of respondents lived in ZCTAs with a high level of WUI, based on this definition, and we classified these respondents as 鈥渓iving near WUI鈥�. The WUI percentage variable was linked to the survey data based on the ZCTA of residence.

Statistical analysis

Weighted logistic regression models were used to evaluate associations between the predictor variables (past experiences with smoke, wildfire risk perception) and the two main outcome variables (support for prescribed burning in general, support for prescribed burning near their home). Models were adjusted for the following potential confounders selected a priori: age, race/ethnicity, household income, sex, and education level. For the models of past experiences with wildfire smoke and wildfire risk perception, we explored the impact of additional adjustment for living near WUI.

Additionally, to assess whether the perceived future risk of wildfires is a mediator on the causal pathway between past wildfire smoke experience and support for prescribed burns, we first used logistic regression to assess whether past wildfire smoke experience is associated with increased risk perception for future wildfires. Next, we applied causal mediation analysis to test this potential pathway. Causal mediation models were adjusted for age, race/ethnicity, household income, sex, and education level, and accounted for the survey weights. All analyses were conducted in SAS statistical software version 9.4 (SAS, Cary, North Carolina).

Results

About 66% of the population expressed support for prescribed burning in general and 54% percent expressed support for prescribed burning occurring near their neighborhoods. Higher percentages of the following groups expressed support for prescribed burns: older adults, non-Hispanic whites, males, people with graduate degrees, people with higher household incomes, and people living near the WUI (Table听1). There were higher levels of support for prescribed burns among people who had experienced wildfire smoke within the past 3 years. Given the sampling strategy and post-stratification weighting based on census data, the weighted distributions shown in Table听1 should approximate the distribution of demographic characteristics of the western U.S. population.

Table 1 Demographic characteristics of survey respondents

Among the subset of respondents who had experienced wildfire smoke within the past 3 years, those who had a higher degree of overall impacts from wildfire smoke in the past 3 years were somewhat more likely to express support for prescribed burning both in general and near their neighborhood, although the confidence intervals overlapped (Fig.听1, panels (a) and (c)). However, respondents who reported a higher degree of health impacts from wildfire smoke were less likely to express support for prescribed burning in general or in the neighborhood, although confidence intervals overlapped (Fig.听1, panels (b) and (d)).

Fig. 1

Support for prescribed burning based on recent experiences with wildfire smoke

Notes: (a) Percent support for prescribed burning in general, based on wildfire smoke impacts overall, (b) Percent support for prescribed burning in general, based on wildfire smoke impacts on health, (c) Percent support for prescribed burning near neighborhood, based on wildfire smoke impacts overall, (d) Percent support for prescribed burning near neighborhood, based on wildfire smoke impacts on health. Error bars display the 95% confidence interval

PB鈥�=鈥塒rescribed Burning

WS鈥�=鈥塛ildfire Smoke

Regarding public interest in receiving more information about prescribed burns (Table听2), 19.3% (95%CI 16.9鈥�21.7) expressed a strong interest in receiving information about how prescribed burns are conducted and 25.5% (95%CI 22.8鈥�28.2%) were interested in information about how prescribed burns reduces the risk of wildfires. About 22.9% (95%CI 20.3鈥�25.6) were interested in learning about what precautions to take while prescribed burns are being conducted, and approximately one-third (33.5%, 95% CI: 30.5-36.5%) expressed a strong interest in receiving information about when and where prescribed burns are being conducted.

Table 2 Estimates of the percent of the population in western conterminous U.S. who expressed a strong interest in receiving specific types of information about prescribed burns

In models adjusting for demographic characteristics (age, race/ethnicity, sex, household income, and education) and WUI status, we found that adults who had experienced a wildfire smoke event recently had about two times the odds of supporting prescribed burning in general compared to people who had not experienced a wildfire smoke event (aOR鈥�=鈥�2.03, 95%CI 1.51鈥�2.74, Fig.听2, panel (a)). The association between recent wildfire smoke experience and support for prescribed burning occurring near their residence also showed a positive association, with those experiencing recent wildfire smoke being 59% more likely to express support for prescribed burning near their neighborhood (aOR鈥�=鈥�1.59, 95%CI 1.20鈥�2.09, Fig.听2, panel (b)), in adjusted models.

Fig. 2

Associations between recent experiences with wildfire smoke, perceived future risk of wildfire smoke, and (a) support for prescribed burning in general and (b) support for prescribed burns near their neighborhood

Notes: The figure displays the odds ratio and 95% confidence intervals in models adjusting for demographics and wildland urban interface or intermix status

WS鈥�=鈥塛ildfire Smoke

Perceiving a high risk of wildfire smoke impacts in the next 12 months was associated with increased support for prescribed burning in general (aOR鈥�=鈥�1.66, 95% CI鈥�=鈥�1.15鈥�2.39) and near their residence (aOR鈥�=鈥�1.72, 95%CI鈥�=鈥�1.23鈥�2.39). Additionally, perceiving a future increased frequency of wildfire smoke events in the next 5鈥�10 years was associated with 52% (95% CI鈥�=鈥�1.12鈥�2.06) and 27% (95%CI鈥�=鈥�0.96鈥�1.68) higher adjusted odds of supporting prescribed burning in general and near a respondent鈥檚 home.

Within the sample of people who indicated they had experienced wildfire smoke in the past 3 years (n鈥�=鈥�708), we examined whether the degree or type of wildfire smoke impact had any influence on their support for prescribed burns (Fig.听3). After adjusting for demographics and WUI status, we found that those who reported experiencing high overall impacts of WS or high impacts on their outdoor air quality were more likely to be supportive of prescribed burning in general when compared to adults who had experienced low impacts (aOR鈥�=鈥�1.52, 95%CI鈥�=鈥�0.75鈥�3.10 for overall impacts, and aOR鈥�=鈥�2.02, 95%CI鈥�=鈥�0.89鈥�4.57 for outdoor air quality impacts), although confidence intervals included the null value. Having high impacts on the ability to engage in physical exercise was positively associated with support for prescribed burning near their residence, although the CI included the null value; this factor was not associated with general support for prescribed burning. High or moderate impacts of WS on a respondent鈥檚 overall mood or physical health were not found to be associated with support for prescribed burning generally nor near their residence.

Fig. 3

Associations between recent wildfire smoke impacts and support for prescribed burning (a) in general and (b) near their residence

Notes: Odds ratios and 95% confidence intervals are from model adjusted for demographics and WUI status. WS鈥�=鈥塛ildfire Smoke

Past WS experience was found to be strongly associated with increased perceived risk for WS events. In models adjusted for demographic variables, those who had experienced WS in the past 3 years had more than 3 times the odds of perceiving a high increased risk of being personally impacted by WS in the next 12 months (aOR鈥�=鈥�3.72, 95%CI鈥�=鈥�2.71鈥�5.13) and increased frequency of WS events in the next 5鈥�10 years (aOR鈥�=鈥�3.17, 95%CI鈥�=鈥�2.40鈥�4.19).

When modeling high perceived risk of being personally impacted by WS in the next 12 months as a potential mediator of the relationship between recent WS experience and support for prescribed burning in general, there was not enough evidence to identify high perceived risk as a significant moderator. However, when assessing high perceived risk as a potential mediator of the relationship between recent WS experience and support for prescribed burning near their residence, the percentage mediated trended towards significance (16.85%, 95% CI=-1.09-34.80, p鈥�=鈥�0.066, Table听3). However, given the very wide confidence interval that includes the null value, this result should be interpreted with caution. Perceiving a high likelihood of an increased frequency in the next 5鈥�10 years was not found to be a mediator between recent WS experience and support for prescribed burning in general or near their residence.

Table 3 Results for high perceived risk of being personally impacted by wildfire smoke in the next 12 months as a mediator between recent wildfire smoke experience and support for prescribed burning near their residence. Models were adjusted for demographic variables

Discussion

Among this sample of adults in the western United States, we found evidence that having recent experience with wildfire smoke, degrees of perception of impact from recent wildfire smoke incidents, and having a high perceived risk for future wildfire smoke incidents are each positively associated with support for prescribed burning. In particular, for those who had recently experienced WS, perceiving a high degree of WS impact on outdoor air pollution was strongly associated with a higher likelihood of supporting prescribed burns, compared to perceived impacts of WS on other aspects of life, such as their health, exercise, and mood. A more proximal perceived risk (within the next year) was more strongly associated with support for prescribed burning compared to perceived risk over a longer time frame (5鈥�10 years). The mediation model data were suggestive that the perceived risk of WS in the next 12 months may be in part building the relationship between recent WS experience and support for prescribed burning near their residence.

This study adds to the literature on factors influencing public support for prescribed burning by providing recent data from residents across a large region, rather than focusing only on smaller, specific high wildfire risk areas. Additionally, by assessing recent experience with wildfire smoke (rather than wildfire risk itself), the study identifies a far more commonly experienced impact of large wildfires and connects these experiences with support for prescribed burning. The timing of this study, following 3 years of very damaging wildfires in this region, captures a unique time frame when there is increased public awareness of wildfire dangers and potentially increased interest in land management tools to help prevent severe wildfires.

A strength of this study is that we are capturing broad public opinion on the topic of support for prescribed burns while also being able to adjust for WUI status, since residents in or near the WUI would likely be the ones most impacted by direct wildfire risk as well as smoke from prescribed burns. Overall, we found that the associations between perceived wildfire smoke risk, recent WS experience and support for prescribed burning were not sensitive to adjustment for WUI status. Therefore, we interpret these results to mean that this potential pathway linking past experiences, perceived risk, and support for risk reduction activities is broadly applicable to both WUI and non-WUI communities.

The trending mediation analysis results suggest a potential strategy for increasing support for prescribed burning by raising awareness of heightened future wildfire risks due to the changing climate. Research shows that climate shifts may narrow the favorable window for prescribed burning, highlighting the need for public support, agency messaging, and continued research into wildfire risk reduction strategies and public motivations to implement them [52]. Our findings showing previous wildfire smoke experience and increased risk perception, each being positively associated with support for fire risk reduction measures is generally consistent with prior studies. A study based in Pend Oreille County, Washington identified perceived risk of wildfires on public lands as an important mediator between perceived risk of human-caused fires on private lands and support for regulations on private lands to mitigate fire risk [53]. A Colorado study of residents near a National Forest found that recent wildfire experience made both prescribed burning and mechanical thinning more acceptable [42]. In a study conducted in Orange County, California, Masri et al. found that previous experience with wildfire was positively associated with support for a tax increase to bolster the firefighter workforce and 鈥渋mprove access to fire safety resources鈥�, although this study did not specifically mention fire prevention strategies [39]. Raising awareness of the wildfire risks and the health impacts of wildfire smoke, along with the benefits of prescribed fire in reducing such risks, may be an important strategy to build public support toward funding these activities. One consideration for future research is a more detailed analysis of the timing between wildfire events and how they may relate to risk perception.

Our findings showing a strong association between greater reported impacts of recent wildfires on outdoor air quality and support for prescribed burns stood in contrast to the finding that greater reported impacts on health showed no association with the outcomes examined. Although these findings may seem disparate, prior research has reported that people who experienced health impacts from previous wildfires are overall less tolerant of smoke from prescribed burns than other adults [54]. In a recent California study, Hoshiko et al. identified that better adaptive capacity among medically vulnerable residents to mitigate the health effects of smoke helped to increase their satisfaction with the information received about prescribed burns [37], highlighting the important role of supporting communities in implementing health protective solutions. Additionally, our findings showing that past impacts of WS on a person鈥檚 mood were not associated with support for prescribed burns suggest that perhaps there is an element of control that may shape these opinions. For example, while there is substantial evidence of WS impacting physical health and some evidence that WS has been shown to have impacts on mental health [55], there are many other factors that could contribute to protecting or improving a person鈥檚 physical health or mood, including adaptive responses such as engaging in indoor social activities [56]. However, prescribed burns would have a direct impact on preserving outdoor air quality when there are few other actions that could accomplish this outcome. The finding that WS impacts on exercise were positively associated with support for nearby prescribed burns but not for prescribed burns in general are surprising, suggesting the contribution of other factors, such as different patterns of engagement in indoor exercise that also affect perceptions of prescribed burns.

In addition, our findings showing strong interest among one third to one fifth of households regarding specific types of information about prescribed burns, particularly safety communications around when and where they are occurring, may represent an opportunity for agencies to conduct outreach to the broader community. Such outreach would help garner broader support for these activities, such as support for increasing the use of prescribed fire or increasing public funding for the workforce needed to conduct prescribed burns. Prior studies identified that trust in the agency conducting the fuel treatment along with understanding the benefits of prescribed burns were critical factors in building public support for this management practice [43, 57]. Additional studies have shown that communication around prescribed burning may be an important element in building this trust. A recent study focused on the mid-Atlantic region of the United States (a relatively lower-risk wildfire region) identified that broader awareness of prescribed burns occurring locally is an important factor influencing public support for the practice [58], also suggesting that improved outreach to nearby communities may help raise awareness and build confidence in the fire agency鈥檚 land management practices. A study in a high wildfire risk area of California found that residents strongly support the use of prescribed fire, but are also very sensitive to the health effects of smoke; residents would benefit from advance notice about when and where to expect smoke, along with guidance related to mitigating smoke impacts on air quality [38]. These public outreach and communication activities could be incorporated into overall land management planning activities, to improve community awareness and support these important risk reduction strategies.

Although our study provides unique information on this topic, there are some key limitations to note. The questionnaire did not ask about alternative strategies for fuel reduction that may be important, such as mechanical thinning or a combination of mechanical thinning and prescribed fire [59,60,61,62]. However, we chose prescribed fire as the focus of this study because of public concerns around the smoke impacts of wildfires, as well as the small risk of escape, which is a particular concern for WUI communities [36,37,38, 42, 63]. We also did not ask specifically about prescribed fire on public versus private lands; the distinction is important in terms of who pays for the fuel treatment activities and the relevant policies that would support such activities [64,65,66]. The findings from the current study are more applicable to public opinion on fuel treatment activities conducted by public agencies, rather than on private lands. While exposure to direct wildfire hazard itself, such as through mandatory evacuations or fire damage to property, was not assessed in the survey, this 鈥渄irect exposure鈥� could also influence perspectives on prescribed burns. Future studies may consider both smoke and fire exposures simultaneously to compare their influence on attitudes toward prescribed fire .

The survey methodology using an online non-probability panel and offering the survey only in English limits the generalizability of the study, with self-selection bias, and response bias among panel members, along with the exclusion of populations with limited English proficiency and those who do not have internet access. Prior research on these panel populations have shown that they are more likely to be younger and have lower incomes compared to average adults in the U.S [67]. Using stratified sampling and standard demographic weighting helps correct some of this bias, as has been employed in the current study [68]. While non-probability panels have been successfully employed in a wide range of social science and public opinion research [69], we nevertheless recognize the limitations and exclusions of this method. Although the sampling method provides an estimate representative of the Western population, the sampling was not designed to allow for state by state comparisons. Future studies may consider a sampling design that produces state-level representation to study differences in messaging across states or EPA regions. Additionally, the use of a single cross-sectional survey results in unknown temporal ordering between the exposure variable (recent WS experience) and the hypothesized mediator (perceived WS risk), which would violate one of the key assumptions in mediation analysis [70]. However, it is a reasonable expectation that the respondents鈥� experience with WS in the preceding 3 years helped to shape their current perception of wildfire smoke risk.

Because there is no standard definition of what constitutes 鈥渓iving near the WUI鈥�, we selected the 40% threshold based on what would be a substantial, recognizable percentage of land area within that zip code. To test whether our results were sensitive to this definition, we conducted a simple sensitivity analysis using an alternative threshold of 50% WUI. Regression models adjusting for WUI status did not differ substantially based on this alternative definition, and the definition did not change the interpretation of the results. The WUI definition used in Carlson et al. 2022 is based on housing density and proximity, which are important contributors to wildfire risk. Other factors such as vegetation density, proximity to industrialized areas, and topography may also play a role; such factors could also contribute to perceived risks and impact perspectives on prescribed burning [71,72,73]. In general, adjustment for WUI status did not change the point estimates or the confidence intervals meaningfully in the logistic regression models, so we did not include the WUI status as an adjustment factor in the mediation analysis. Finally, while our study results are indicative of the associations described, results with very wide confidence intervals and/or borderline significance must be interpreted with caution. Replication studies are needed to confirm these results and to provide additional insight.

Conclusions

Our study showed that recent experience with wildfire smoke and risk perception of future smoke events are positively associated with support for prescribed burning. These findings suggest that upcoming efforts to help build understanding and shape public sentiment in the U.S. western region are important as land managers work toward expanding the pace and scale of fuel treatments to reduce wildfire risk. Opportunities for improving community outreach and developing effective public notification strategies would help keep the public informed about the occurrence of prescribed burns near their neighborhood, provide credible information about climate change and wildfire risk, build trust in land management entities, and provide information about health-protective actions.