Voluntary uptake of flood insurance in post-fire landscapes: insights from household surveys after two wildfires in northern Arizona

Wildfires can rapidly generate post-fire flood risk for downslope communities, complicating recovery and necessitating accelerated cross-boundary responses on both public and private lands. Flood insurance is one of a suite of efforts to minimize the social and financial impacts associated with post-fire flood events, but little is known about who purchases coverage after wildfires or their experiences with insurance claims. We provide an opportunistic contribution to this research gap by comparing data gathered via three mixed-mode household surveys conducted in areas with modeled flood risk after two wildfires in the greater Flagstaff, Arizona area. Two surveys were administered—one in 2011 and one in 2020—after the 2010 Schultz Fire, which resulted in significant flooding, providing a longitudinal insight into insurance experiences. A third survey was administered in 2022 following flooding adjacent to the 2019 Museum Fire. Together, these surveys resulted in a total of 1,351 usable responses. We find that several groups are significantly less likely to purchase flood insurance, including residents who moved to their property after the studied fire event and renters. The reported cost of flood insurance among households affected by the Schultz Fire doubled over a 10-year period; those who no longer maintained coverage typically reported that they stopped renewing between 2–5 years after the fire. Second homeowners were more likely to report damages that were not covered by insurance. We provide recommendations for engaging residents in uptake, renewal, and use of flood insurance and conclude that there is an urgent need to better understand decision-making surrounding post-fire flood insurance to support more equitable recovery.


Introduction
Wildfires remove vegetation and can cause soils to become hydrophobic, transforming typical rainfall events into flash floods (Shakesby and Doerr 2006).Downslope communities must act rapidly to address this emergent risk in the often short and unpredictable window between a wildfire's ignition and the first forecast precipitation -a challenge that requires substantive communication, coordination, and resources to minimize damages across jurisdictional boundaries (Youberg et al 2019, Burnett andEdgeley 2023).Purchase of flood insurance by private property owners within this post-fire window is commonly recommended as one action in a suite of approaches to mitigate potential impacts as property damage or loss associated with wildfires rise (FEMA 2020, Buechi et al 2021).Although flood insurance is mandatory in high-risk floodplains across the US, most post-fire flooding emerges in places that were not previously documented as flood-prone, which may leave households unexpectedly vulnerable after wildfires (Kousky and Michel-Kerjan 2017, Serra-Llobet et al 2023).Residents seeking flood insurance coverage after wildfire may voluntarily purchase coverage from federal programs like the National Flood Insurance Program (NFIP) or through private companies (Kousky and Shabman 2014).Little is known about who purchases flood insurance in post-fire environments, how long they renew coverage, or the experiences those households have making claims associated with post-fire flood impacts (Eriksen and de Vet 2021, Burnett andEdgeley 2023).The costs of these impacts are diverse and often go unaccounted in traditional estimates of fire and flooding costs where suppression expenditure is prioritized (Hjerpe et al 2023).Understanding household behaviors towards post-fire flooding is critical to advance social adaptation to wildfire and related cascading consequences, particularly in landscapes across the Southwestern US where monsoonal rainfall can cause extensive damage and loss of life (Martin et al 2007, Paveglio et al 2019).
Existing research that seeks to document factors motivating uptake of flood insurance predominantly entails household surveys or secondary data analysis of households in coastal and floodplain areas.Sociodemographic characteristics are often related to uptake of insurance coverage, including age, level of education, income, and ethnicity, but their significance varies across contexts (Landry and Jahan-Parvar 2011, Atreya et al 2015, Shao et al 2017, Cannon et al 2020).Other considerations include home ownership, trust in government, social expectations, and eligibility for disaster assistance (Lo 2013, Petrolia et al 2013, Shao et al 2017, Zinda et al 2021).Additionally, the broader hazards literature consistently identifies perceptions of risk and vulnerability as motivators of insurance purchasing after both flooding and wildfire (Martin et al 2007, Petrolia et al 2013).Recent flood events can cause upticks in purchase of insurance coverage that typically decline at the three year mark (Atreya et al 2015).The presence or absence of insurance coverage has cascading consequences during recovery in instances where flood or fire events do occur, creating or exacerbating inequities within at-risk communities (Kousky 2019).This includes the type or extent of coverage, as recent studies find high rates of underinsurance, particularly in rural areas (Osbaldison et al 2018, Booth et al 2022).Clarifying who does and does not purchase voluntary flood insurance after wildfires can help governments anticipate potential needs and more intentionally target specific groups with communication about the value of insurance coverage.
While the flood and wildfire insurance literature investigating related human behaviors and decisionmaking is growing, few efforts have been made to understand whether they translate to post-fire flood environments, or the extent to which those behaviors vary across populations following different wildfires.Such research is central to future risk communication and recovery strategies in places like the western US where social understandings of post-fire flooding are in demand (Edgeley andColavito 2022, Edgeley 2023).This research note presents an opportunistic analysis of three surveys in post-fire landscapes with the intent to establish a foundation for future research, focusing on the following research questions: 1. To what extent do residents engage with flood insurance after wildfire?2. What factors characterize uptake of flood insurance and its use after wildfire?

Study fires
The city of Flagstaff, Arizona, and the surrounding area has recently experienced several wildfires that produced post-fire flood risk.Two prominent human-caused fires ignited on the Coconino National Forest-the 2010 Schultz Fire caused by an abandoned campfire and the 2019 Museum Fire believed to have been ignited by an isolated spark from contractor equipment on a fuels management project-generated flood risk for thousands of homes in several downslope neighborhoods both in and outside of city limits.The Schultz Fire burned 15,075 acres, causing frequent flood events driven by monsoonal rains over the burn scar for several years (Combrink et al 2013).This post-fire flooding caused one fatality and extensive damage to private property, resulting in total estimated costs between $109 and $114 million USD (Hjerpe et al 2023).In 2010, purchase of flood insurance entailed a 30-day wait period before coverage began, meaning that many residents had damages that were not yet covered under their policies.The Museum Fire burned 1,961 acres in close proximity to neighborhoods, causing it to become the highest suppression priority in the US at the time (Colavito et al 2023).Following the fire, concerns about post-fire flooding mounted but the summers of both 2019 and 2020 were characterized by drought conditions and unusually low precipitation.In the summer of 2021, monsoon rains returned, and subsequent flooding was produced both from the burn scar and unburned areas, converging on a cluster of neighborhoods and causing widespread damage.In this survey, addresses both in and out of the modeled burn scar flood risk area were surveyed.Both fires presented unique opportunities for addressing research gaps related to post-fire flooding, insurance coverage, and the recovery process after wildfire (Edgeley 2023).

Survey methodology
Household surveys were conducted in areas identified as at risk of post-fire flooding after the 2010 Schultz Fire and the 2019 Museum Fire (figure 1).Two surveys were administered for each fire: the Schultz Fire in 2011 and 2020, and the Museum Fire in 2019 and 2022.Given that flooding had not yet occurred after the Museum Fire in 2019, this study focuses on the 2011, 2020, and 2022 surveys (see Edgeley and Colavito 2022 for 2019 survey findings).Both the 2011 and 2020 Schultz Fire survey sample frames focused on areas where flooding was projected to occur; however, new models projecting 100-year flood risk were released between studies meaning that the 2020 survey sample frame covered a far greater area.The Museum Fire 2022 survey sampled all residents south of the burn scar and north of Route 66, which included post-fire flood risk depths of two inches in 45 minutes modelled in 2022 (Loverich 2022) as well as adjacent households without modeled flood risk and households affected by recent flooding not associated with the burn scar.As a result, some households without post-fire flood risk were surveyed; however, some of those households still purchased flood insurance out of caution or because they perceived post-fire flood risk to their property so their responses are included in this study.Response rates varied across surveys; the 2011 Schultz survey received 321 responses for a 24% response rate with the 2020 follow up gathering 407 responses for a 22.6% response rate, while the 2022 Museum Fire survey had 623 responses for a response rate of 16.3%, providing a total of 1,351 completed questionnaires for this analysis.
Survey instruments across all three studies shared commonalities in the following question topics: (1) flooding experiences and damages caused by flood events, (2) insurance purchase, renewal, and claims, (3) property mitigation efforts and related costs, and (4) in the case of the 2020 and 2022 surveys, basic sociodemographic information.In each study, questions were intentionally designed to encourage responses from both residents who lived in the area prior to the study fire and those who had moved into the area since.Questions used here are primarily binary, multiple choice, or required the estimation of dollar amounts.Survey instruments and additional methodological context are available in Combrink et al (2013), Colavito et al (2021, 2023), and Edgeley and Colavito (2020, 2022).
In each study, survey materials were mailed to households in several phases: 1) a survey packet containing an introductory letter with information about the study, a questionnaire booklet, and a pre-paid return envelope; 2) a reminder letter or postcard one to two weeks later; and 3) in the case of the 2020 and 2022 studies, a final reminder letter after an additional two weeks to incentivize responses from those yet to participate following Dillman et al's (2014) guidance.Respondents also had the option to complete the 2020 and 2022 surveys online via a link provided in each mail out to maximize opportunities to participate (Börkan 2010).
Exploration of the resulting data was conducted in statistical analysis software SPSS using descriptive statistics, two-sided Fishers exact tests, chi square tests with a Bonferroni correction, and ANOVA.These tests allowed for comparisons both between surveys and across all respondents to determine statistically significant differences in factors influencing purchase and use of insurance (Field 2018).The findings presented below draw upon data from questions focused on insurance in each of the three surveys described above and expand upon existing reporting to add a comparative analysis across the studies that provides greater depth and context for emerging understandings of relationships between post-fire flooding and insurance uptake

Flood insurance coverage and use after wildfire
Table 1 reviews responses to insurance questions across all three surveys.Purchase and retention of flood insurance coverage varied significantly across all three surveys.Half (50%) of 2011 Schultz survey respondents held flood insurance a year after the first 2010 flood event compared to 22.3% of Museum Fire respondents a year after the 2021 flooding.A decline in coverage occurred for the Schultz Fire between the 2011 survey (50%) and the 2020 survey (39.6%), while rate of annual renewals decreased 40.6%.Respondents to the 2020 Schultz survey who had purchased flood insurance after the fire but no longer renewed it as of 2020 (n = 86) were asked which year they first cancelled their coverage; 67.4% stopped renewing at between two and five years after the fire (2012-2015), with a peak of 24.4% dropping coverage in 2012, two years after the fire event.Similarly, Museum Fire respondents who had purchased flood insurance were asked how long they anticipated renewing their coverage; 27.8% expected to stop renewing within the next five years, but 50.0%expected to renew for the duration of their tenure at the surveyed address.Intended renewal of insurance for longer periods of time was motivated by perceived likelihood of a flood event impacting the respondent's property within the next 10 years (p = .027).Schultz Fire survey recipients were asked to share the type(s) of coverage their insurance provided, revealing a small decrease in home contents, structure, and renter's insurance coverage while coverage of outbuildings like garages increased.Schultz Fire survey respondents reported that the cost of their coverage had nearly doubled from an average rate of $357 in 2011 to $613 in 2020.
Between 30% and 45% of respondents to each survey reported damage to their property because of flood events after wildfire.American Indian or Alaska Native (p = .025)and Asian (p = .003)respondents were more likely to report flooding on their property.The most common damage across all surveys was non-structural (e.g., landscaping or driveway damage), while the least common were damages to structure foundations and interiors.Respondents to the 2011 Schultz survey were more likely to report damages than the subsequent two surveys, which could indicate higher levels of turnover in property ownership for the Schultz Fire study area.A small subset of Schultz and Museum Fire respondents made insurance claims related to flood damages.The average estimated cost of damages increased substantially between surveys, most notably between the 2011 and 2020 Schultz surveys, demonstrating the cumulative nature of flood damage expenses over time.Survey respondents appeared underinsured after both fires; Schultz Fire respondents received average payouts covering approximately 55.9% of the total estimated damage cost, while Museum Fire respondents received payouts covering 28.2%.Around 22.5% of 2020 Schultz survey respondents and 15.6% of 2022 Museum Fire respondents reported damages or expenses that were not insured, typically described by respondents as debris removal, landscaping damage, loss of pasture for livestock, or interior structure flooding.In 14 instances following the Schultz Fire, respondents were unable to make a claim because the damages occurred within the 30-day period before flood insurance came into effect.The cost of uninsured damages averaged $12,111 after the Schultz Fire and $33,362 after the Museum Fire.Repairs associated with damage from flooding after the Schultz Fire took an average of 26.9 months to complete according to 2020 Schultz survey respondents; 10.5% of those with damages stated that repairs were still in progress almost 10 years after the fire.
Respondents to the Schultz Fire surveys were asked to provide information on the cost of mitigation efforts that were not covered by insurance (e.g., waterproofing residential structures, improving property drainage).The average out of pocket costs for mitigation rose from $3,089 in 2011 to $7,227 in 2020.Respondents in 2020 were also asked whether out-of-pocket costs associated with the upkeep of these mitigations had emerged; 37 respondents reported an average upkeep cost of $3,620.Together, these responses suggest that the average household spent around $10,847 in post-fire flood mitigation in the ten years following a wildfire.

Characterizing flood insurance buyers after wildfire
Several factors influenced the purchase of flood insurance.The most significant of these was whether the respondent's property had been damaged by flooding (p = <.001);these respondents were also more likely to renew their coverage annually (p = <.001).Those who had moved into their surveyed property after a study fire were more likely to report being aware of potential flood risk when they moved into their property (p = <.001);however, they were also significantly less likely to obtain flood insurance (p = <.001) and renew it (p = .005),and also significantly less likely to file an insurance claim for flood damages (p = .006).Renters were also significantly less likely to obtain or renew flood insurance after fires compared to property owners (p = .002);however, respondents who owned surveyed rental properties were far more likely to purchase coverage for that unit than owner-occupied properties (p = .015)and renew it annually (p = .018).Part time residents were more likely to report impacts for which they did not file an insurance claim (p = .049).Gender, education, age, ethnicity, and annual household income had no impact on purchase or renewal of flood insurance.
Respondents to the 2020 Schultz and 2022 Museum Fire surveys were asked about their level of agreement with the statement 'my property is insured, so mitigation actions seem less necessary.'American Indian or Alaskan Native respondents (p = .010),Hispanic or Latino respondents (p = .002),younger respondents (p = .034),and respondents from households with higher income (p = .003)were significantly more likely to agree.Museum Fire survey recipients were also asked about their risk perceptions.The higher the perceived risk, the more likely the respondent was to have flood insurance (p = <.001) and renew it annually (p = <.001).

Discussion
This research note responds to calls for a greater understanding surrounding flood insurance coverage and use after wildfire with the intent to provide a foundational understanding of both the factors influencing purchase and how uptake of insurance in post-fire spaces may differ from traditional flood insurance studies.Data collection to determine household impacts associated with post-fire flooding is rare, particularly regarding undocumented expenses and the role of insurance in household recovery (Hjerpe et al 2023).We find that postfire flooding insurance behaviors closely align with behaviors identified in studies of wildfire and flood insurance outside of post-fire environments; however, there may be differences in household experiences that vary with fire location and impact that can be identified and explored with additional studies in other states or countries and with the use of qualitative research methods.
Flood insurance coverage appears to decline over time post-fire in all three surveys.Several factors may be influencing this decline, including increasing costs associated with insurance coverage, underinsurance, out-ofpocket costs associated with flood risk mitigation, and decreasing risk perceptions.We found that households most consistently stop renewing between two to five years after a fire; therefore, renewed or increased outreach efforts regarding uptake and maintenance of insurance coverage at the two-year mark may be important in areas where post-flood risk is still present.Some private insurance companies incentivize wildfire mitigation by offering discounted rates for participants in formalized risk mitigation programs like Firewise, indicating that a similar framework for post-fire flood mitigations on private property could generate more affordable premiums that lead to more continuous coverage.Underinsurance was also common across both fires-a pattern that is increasingly present after hazard events and that reiterates the importance of multi-faceted risk adaptation rather than dependence on insurance alone.However, out-of-pocket mitigation and upkeep costs were significant, suggesting that development of grants or cost-shares for these efforts may generate less economic inequity surrounding risk reduction.The longitudinal nature of the Schultz Fire surveys indicated that even at the 10-year mark many still retain flood insurance after fire, but our data suggest that without the occurrence of flood events that provide 'evidence' of the need for coverage, it is unlikely that insurance purchase and renewal would remain as high.
Inequities regarding both insurance coverage and post-fire flood impacts were present across surveys.Respondents who identified as ethnic or racial minorities, particularly those who were American Indian or Alaska Native, Asian, and Hispanic or Latino, appeared to be more vulnerable to post-fire flood impacts, and were more likely to feel that flood insurance negated the need for mitigation efforts on their property.This suggests that communication about sandbag use, flood event safety, and other mitigation activities could benefit from targeting minority groups to reduce their vulnerability after post-fire flooding.Renters' vulnerability appears heavily dependent on the willingness of the property owner they are leasing from to purchase coverage and take action.While in our study, respondents identifying as rental property owners appeared engaged and willing to invest in post-fire flood protection, research tied to other hazards indicates that this is not always the case (Heiman 2021).Lastly, residents who move into post-fire spaces are most likely to report being aware of flood risk upon arrival yet least likely to acquire flood insurance; we suggest that this disconnect should be examined in greater detail and may benefit from neighbor-to-neighbor discussions of risk, communications from realtors or local government, or targeted outreach programs by trusted organizations to encourage uptake of coverage.
This preliminary assessment of households in areas with post-fire flood risk invites additional studies that can more comprehensively explore risk adaptation and the role of insurance.Future research should seek to (1) explore differences in experience with federal NFIP coverage and private coverage, (2) examine the causes of underinsurance and strategies to address them, (3) understand the impact that insurance coverage or lack thereof can have on broader household finances and wellbeing, and (4) combine hydrological studies of enduring post-fire flood risk with long-term retention of insurance coverage.Tensions surrounding insurance coverage associated with wildfire risk have motivated the generation of state level policies like California's Fair Access to Insurance Requirements (FAIR) plan intended to address areas no longer supported by private companies.As post-fire flood risk continues to emerge with recent upticks in wildfire occurrence, size, and severity, proactive efforts to examine flood insurance coverage in these spaces can inform more equitable and comprehensive protection.

Conclusion
Flood insurance is one of a suite of efforts to minimize the social and financial impacts associated with post-fire flood events.Findings presented here provide an opportunistic contribution to research gaps surrounding understandings of flood insurance-related behaviors in post-fire environments.Minority groups appear to be less engaged in the uptake of insurance, indicating that recovery from a post-fire flood might generate significant inequities in repair or rebuilding on private property.Other groups such as renters who are often subject to the insurance decisions of the rental property owner are also vulnerable.The reported cost of flood insurance among households affected by the Schultz Fire doubled over a 10-year period; those who no longer maintained coverage typically reported that they stopped renewing between 2-5 years after the fire.These findings, among others, indicate that communication and outreach intended to encourage the uptake of flood insurance after wildfires must be multi-faceted and nimble to demonstrate value to these different groups, whether that be the development of content in multiple languages or initiatives targeted at different points in time to align with resident knowledge and risk level.We encourage continued study of flood insurance behaviors and experiences after fire in locations beyond northern Arizona, particularly in areas where repeat or longitudinal studies may be feasible, in order to characterize how behaviors evolve relative to changing post-fire flood risk.

Figure 1
Figure 1 Map of the 2010 Schultz Fire and 2019 Museum Fire Perimeters and 2020 and 2022 Survey Sample Areas in the Greater Flagstaff, AZ Area.The 2011 survey sample entailed a subset of addresses within the 2020 survey sample area.

Börkan B 2010
The mode effect in mixed-mode surveys: mail and web surveys Soc.Sci.Comput.Rev. 28 371-80 Buechi H, Weber P, Heard S, Cameron D and Plantinga A J 2021 Long-term trends in wildfire damages in California International Journal of Wildland Fire 30 757-62 Burnett J T and Edgeley C M 2023 Factors influencing flood risk mitigation after wildfire: insights for individual and collective action after the 2010 Schultz Fire Int.J. Disaster Risk Reduct.94 103791 Cannon C, Gotham K F, Lauve-Moon K and Powers B 2020 The climate change double whammy: flood damage and the determinants of flood insurance coverage, the case of post-katrina new orleans Climate Risk Management 27 100210 Colavito M M, Combrink T, Hjerpe E, Edgeley C, Burnett J and Sánchez Meador A 2021 Full-cost accounting remeasurement of the 2010 schultz fire: understanding the long-term socio-economic implications of high-severity wildfire and post-wildfire flooding Ecological Restoration Institute, ERI White paper-Issues in Forest Restoration 45 Colavito M M, Edgeley C M and vonHedemann N 2023 Public experiences with wildfire and flooding: a case study of the 2019 museum fire near flagstaff, Arizona Ecological Restoration Institute, ERI White paper-Issues in Forest Restoration 57 Combrink T, Cothran C, Fox W, Peterson J and Snider G 2013 A full cost accounting of the 2010 schultz fire Ecological Restoration Institute, ERI-Issues in Forest Restoration 44 Dillman D A, Smyth J D and Christian L M 2014 Internet, Phone, Mail, and Mixed-Mode Surveys: The Tailored Design Method (Wiley) Edgeley C M 2023 Social science to advance wildfire adaptation in the southwestern United States: a review and future research directions International Journal of Wildland Fire 32 1647-62 Edgeley C M and Colavito M M 2020 Local experiences with the 2019 museum fire and associated flood risk: a survey of flagstaff-area residents Ecological Restoration Institute, ERI White paper-issues in Forest Restoration 40 Edgeley C M and Colavito M M 2022 Characterizing divergent experiences with the same wildfire: insights from a survey of households in evacuation, postfire flood risk, and unaffected areas after the 2019 museum fire Journal of Forestry 120 660-75 Eriksen C and de Vet E 2021 Untangling insurance, rebuilding, and wellbeing in bushfire recovery Geographical Research 59 228-41 FEMA 2020 Flood Risks Increase After Fires (p. 2) [FEMA Flood After Fire Factsheet] (Federal Emergency Management Agency) https://fema.gov/sites/default/files/documents/fema_flood-after-fire_factsheet_nov20.pdfField A 2018 Discovering Statistics Using IBM SPSS Statistics (SAGE Publications) Heiman E.R 2021 Protecting renters from flood loss U. Pa.L. Rev. 170 783-809 Hjerpe E E, Colavito M M, Edgeley C M, Burnett J T, Combrink T, Vosick D and Sánchez Meador A 2023 Measuring the long-term costs of uncharacteristic wildfire: a case study of the 2010 schultz fire in Northern Arizona International Journal of Wildland Fire 32 1474-86 Kousky C 2019 The role of natural disaster insurance in recovery and risk reduction Annual Review of Resource Economics 11 399-418 Kousky C and Michel-Kerjan E 2017 Examining flood insurance claims in the united states: six key findings Journal of Risk and Insurance 84 819-50 Kousky C and Shabman L 2014 Pricing Flood Insurance: How and Why the NFIP Differs from a Private Insurance Company Resources for the Future Discussion Paper 14-37 Landry C E and Jahan-Parvar M R 2011 Flood insurance coverage in the coastal zone Journal of Risk and Insurance 78 361-88 Lo A Y 2013 The likelihood of having flood insurance increases with social expectations Area 45 70-6 Loverich J 2022 Spruce watershed -pre-fire vs. post-fire hydrologic comparison Memorandum letter to L. Andreani, Public Works Director, Coconino County.Provided by J.E. Fuller, Hydrology and Geomorphology, Inc Martin I M, Bender H and Raish C 2007 What Motivates Individuals to Protect Themselves from Risks: The Case of Wildland Fires Risk Analysis: An International Journal 27 887-900 Osbaldison D N, McShane D C and Oleszek R 2018 Underinsurance in cyclone and flood environments: a case study in Cairns, Queensland Australian Journal of Emergency Management 34 41-7 Paveglio T B, Edgeley C M, Carroll M S, Billings M and Stasiewicz A M 2019 Exploring the influence of local social context on strategies for achieving fire adapted communities Fire 2 26 Petrolia D R, Landry C E and Coble K H 2013 Risk preferences, risk perceptions, and flood insurance Land Economics 89 227-45 Serra-Llobet A, Radke J, Kondolf G M, Gurrola L, Rogers J D, Lindbergh S and Douvinet J 2023 Risk as a process: a history informed hazard planning approach applied to the 2018 post-fire debris flows, Montecito, California Frontiers in Environmental Science 11 24 https:// frontiersin.org/articles/10.3389/fenvs.2023.1183324Shakesby R A and Doerr S H 2006 Wildfire as a hydrological and geomorphological agent Earth Sci.Rev. 74 269-307 Shao W, Xian S, Lin N, Kunreuther H, Jackson N and Goidel K 2017 Understanding the effects of past flood events and perceived and estimated flood risks on individuals' voluntary flood insurance purchase behavior Water Res.108 391-400 Youberg A M, Loverich J B, Kellogg M J and Fuller J E 2019 Before the fire: assessing post-wildfire flooding and debris-flow hazards for predisaster mitigation Natural Hazards and Earth System Sciences Discussions 1-21 Zinda J A, Williams L B, Kay D L and Alexander S M 2021 Flood risk perception and responses among urban residents in the northeastern United States Int.J. Disaster Risk Reduct.64 102528

Table 1 .
Descriptive statistics for key questions about insurance coverage and claims across three surveys following the 2010 Schultz and 2019 Museum Fires in the greater Flagstaff, Arizona area.* Indicates responses from the entire survey sample rather than only respondents who had purchased insurance.