https://orcid.org/0000-0002-7709-1508
Figures
Abstract
Wildfire-caused damage to highways has a significant financial cost to the local, regional, and state jurisdictions where they occur. This study examines the financial ramifications of the harm caused to highways by megafires, using a case study of the highways impacted in the U.S. state of Oregon by the five megafires that occurred during the 2020 Labor Day wildfires. This study proposes a method to classify financial road damage from these wildfires based upon curated datasets collected from the Oregon Department of Transportation (ODOT). Hence, this study presents a dataset with labeled classes, which include physical, roadway, and traffic. Physical consequences included an estimated total temporary and permanent repair cost of $44,894,471, an average permanent repair cost per km of highway affected of $51,705, and an increase of 11% in distance and 11% in time required while using detours. Roadway financial impacts involved around $32,680 per km of highway for hazard tree removal emergency repairs and a decrease of about 14% in the annual average daily traffic (AADT) because of traffic impacts. This paper expands the existing body of knowledge by providing a single source for statistical data required to conduct reliable financial analysis on damages to roadways due to megafires.
Citation: Christiansen K, Mostafiz RB, Al Assi A, Rohli RV, Friedland CJ (2024) Financial impacts of 2020 Labor Day wildfires to Oregon highways. PLOS Clim 3(10): e0000489. https://doi.org/10.1371/journal.pclm.0000489
Editor: Ahmed Kenawy, Mansoura University, EGYPT
Received: December 19, 2023; Accepted: August 24, 2024; Published: October 4, 2024
Copyright: © 2024 Christiansen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All the data used in this manuscript are provided in the appendix. The appendix includes detailed datasets, metadata, and supporting information necessary to replicate the study's findings.
Funding: The authors received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Wildfires inflict millions of dollars of damage to roads every year. While valuable information surrounding the overall cost of wildfires exists [1–3], there is a clear lack of information pertaining to the financial costs associated with roadway damage caused by wildfires [4]. The 2020 Oregon Labor Day Wildfires burned over a million acres (200,000 hectares) [5–7], resulting in significant costs the state had to incur to repair and reopen critical roadway infrastructure.
According to protocol, the federal government is responsible for responding to wildfires that start on federal lands, while individual states manage fires that start on state, local, and private nonfederal lands (aside from those protected by federal agencies under cooperative agreements) [8]. The National Interagency Fire Center, the annual National Fire Plan budget, and federal land management organizations are the main information sources providing wildfire analysis at the federal level [9]. A report entitled, “Strategies for Funding Wildfire Mitigation” describes some of the largest financial loss events in the world, offering strategies to manage the resulting damage effectively [10].
Several methods are utilized to document and analyze the impact of wildfires. From a high-level perspective, the damage caused by wildfires is typically classified as direct, indirect, or post-fire damage [4]. More specifically, a study conducted by the Wildland Fire Lessons Learned Center established eleven distinct ledger categories for categorizing the costs and losses associated with wildfires, including (a) suppression costs, (b) property, (c) public health, (d) vegetation, (e) wildlife, (f) water, (g) air and atmospheric effects, (h) soil related effects, (i) recreation and aesthetics, (j) energy and (k) heritage [11].
Despite the various classification and analysis structures used to evaluate wildfire roadway damage, there is one obvious omission–the financial costs associated with the damage to road infrastructure. Roads are essential assets to our economic systems, yet they are also incredibly susceptible to wildfire damage. If we do not continue to study the risks and damage associated with wildfire damage to roadways or the financial losses likely to result, we will not be able to ensure the future dependability of our transportation systems. One of the major influences of roadways’ increased vulnerability to wildfire damage is climate change, though existing research has not predicted how climate change will impact that susceptibility over time. The study of climate change and its impact on roadways is crucial to long-term planning, strategic investment, and the development of resilience strategies [12] that can influence methods for assessing the financial costs of wildfire roadway damage and developing cost-effective mitigation strategies.
Current literature indicates that the amount of fuel, the hydro-geological and thermo-physical properties of the soils, as well as the intensity, length, and recurrence of fire, all have a significant impact on the thermal soil disturbance that occurs after it ends, potentially causing significant financial expense long after a wildfire has occurred [13].
The contemporary discourse frequently refers to financial difficulties caused by past and present fire management policies in light of climate change and growing human development in fire-prone areas, which includes financial accountability and risk management techniques [14]. The safety, dependability, efficiency, and sustainability of every transportation system—specifically roads—are seriously threatened by climate change and extreme weather events [15]. A congressional study on federal assistance for wildfire response and recovery [16] describes federal options for assistance throughout the life of a wildfire. Strategies being developed to use climate policy to save, restore, and strengthen forests’ ability to withstand the effects of an increasingly disruptive climate include a new financing and responsibility system for climate change which may make longer-term investments in resilience and forest restoration possible. The main components include: (a) strict liability for all greenhouse gas emissions from unintentional fires on federal public lands; (b) requirement that the federal government pay the social cost of carbon for these emissions; and (c) a special fund that would receive these payments and be dedicated to forest restoration, with a requirement that the funds be spent on actual on-the-ground restoration work [17]. Significant research has also been conducted concerning the economic damage of wildfires including property [18], stakeholders/liability, animals, smoke/health, and tourism [19].
A case study that calculates a cost for a sample of people exposed to smoke from the 2009 Station Fire in California serves as an outline of a methodology that provides an empirical technique to measure the financial cost of health impacts linked to exposure to wildfire smoke, which can be applied to damage assessments of future wildfires [20]. It was projected that 10.3 million people in the U.S. suffered from hazardous air quality levels linked to wildfire exposure for longer than ten days between 2008 and 2012 [21,22]. A nationwide assessment of the health effects and medical expenses associated with wildfire smoke exposure was conducted using a new dataset of smoke plumes from wildfires that utilize the location and motions of each wildfire smoke plume in the U.S. and are established by an operational team of experts from the National Oceanic and Atmospheric Administration (NOAA) using satellite imagery. They are used to determine the daily smoke exposure status for nearly all U.S. locations. These data are then combined with daily health and healthcare cost data obtained from administrative records for the whole population of Medicare beneficiaries from 2005 to 2013 to investigate the health impacts of smoke exposure [23].
Research focusing on economics, suppression efforts, and health effects from wildfires solely in Oregon has been conducted as well. A regional social accounting matrix (SAM) model that evaluates the regional economic repercussions from rangeland wildfires was connected in studies conducted in southeastern Oregon on the ranch level linear programming (LP) model [24]. A financial study conducted in eastern Oregon compared fire models with no fire models to assess the risk of an increase in cheatgrass-related wildfires in the Great Basin sagebrush steppe. The study concluded that if the effects of fire on ranches on BLM land are not considered, the costs or benefits of additional invasion will be overestimated [25]. Research conducted in central Oregon’s Deschutes National Forest developed a method for calculating potential wildfire suppression cost savings using a regression cost model and quantified the effects of fuel treatments on wildfire size and suppression costs [26]. The Biscuit fire in southwestern Oregon examined how the fire affected the Douglas-fir log markets over the years since the fire occurred in 2002 [27]. While the existing discourse examines the concepts concerning regional economic models, suppression practices, and costs, they fail to address an accessible process for determining the financial costs associated with roadway damage caused by wildfires or a methodology for utilizing that data to plan and prepare for future damage.
The wildland-urban interface (WUI) has been researched extensively as it relates to the rising costs of preventing wildfires, how they have increased over time, and the two primary causes of the rising costs, which are the increase in biomass fuels and climate change [28]. Residents in the WUI in California may be familiar with the risk of wildfire from forests due to large and frequent wildfires, and they may also be able to compare various policy measures used to support local efforts to create defensible space [29]. While an older source [30] examined the effects of wildfires on local employment and wages in the U.S., a case study conducted around Trinity, California, examined the labor market, suppression spending, and qualitative interview data surrounding the economic impacts of the 2008 wildfires in the area [31]. By studying the increasing severity and intensity of flames as well as the rising percentage of state and federal agency funds allocated to managing fires, research on wildfires in central Oregon concentrated on WUI and the changing climate conditions [32]. The WUI, which accounts for 10% of the land in the U.S., is a significant source of fires that are usually sparked by humans, according to research released in 2020 [33]. The 2020 Oregon Labor Day fires, notably the Beachie Creek and Lionshead wildfires, were investigated in Oregon’s Santiam Basin to explore the wildfire resilience and recovery lesson acquired from them [34]. The National Fire Plan 2000, Federal Wildland Fire Management Policy 2001, 10-Year Comprehensive Strategy 2001, and the Healthy Forests Restoration Act 2003 all identify the WUI fire problem as a major issue and have been in effect since 2000 [35]. Again, the literature above details information surrounding costs associated with suppression costs, climate change, and the WUI, but fails to address the financial damage to roadways due to wildfires.
The approach to this research is to: a) determine how much financial damage the megafires caused to Oregon highways during the 2020 Labor Day wildfires; b) evaluate the location of the highways within the megafires as an indicator of the level of financial damage; c) assess future financial damage to highways based upon historical megafire data. The straightforward analysis of the roadway financial damage acquired from visual inspections obtained from the Oregon Department of Transportation (ODOT) Detailed Damage Inspection Reports (DDIRs) will facilitate identifying consistent ways to measure the cost of roadway damage based upon the size of the wildfire and linear distance (km) of roadway impacted. The suggested approaches and Oregon results are suitable to aid local, state, and federal agencies in formulating sensible budgetary and emergency recommendations for compensatory financial infrastructure against wildfire damage. In addition, results can help agencies to better classify financial roadway damage, which can develop better guidance using the visual–windshield process more effectively to evaluate the potential cost to repair the damaged roadways.
The research contributions include a) creating the categories of financial impacts for highway infrastructure that has not been previously identified; b) establishing proportional statistics based upon the historical financial impact data; c) quantifying the level of financial impacts for multiple categories for the 2020 Labor Day megafires in Oregon.
1.1 Study area
The U.S. State of Oregon highway system was selected for this analysis due to the state’s continued vulnerability to wildfire threats as well as the opportunity to provide insightful information to its state hazard mitigation plan. While wildfires pose a severe threat in Oregon, other dangers such as high winds in the summer and autumn, extreme heat, and dryness can exacerbate and magnify the threat. This analysis focuses on highways that were impacted for several weeks and represents five of the eight Oregon Natural Hazards Mitigation Plan (ONHMP) [36] risk assessment regions. By analyzing the financial impacts to the roads inside the ONHMP natural hazard regions, each of these highways was evaluated at a regional scale and divided into three categories. The routes that were researched, along with the wildfires that touched each of them, are listed below—depicted in Fig 1 and summarized in Table 1.
a. Location of 2020 Labor Day wildfires in Oregon [37]; b. Location of highway 224 within the 2020 Riverside wildfire [37]; c. Location of highway 22 within the 2020 Beachie Creek wildfire [37]; d. Location of highway 22 within the 2020 Lionshead wildfire [37]; e. Location of highway 126 within the 2020 Holiday Farm wildfire [37]; f. Location of highway 138, within the 2020 Archie Creek wildfire [37].
2. Materials and methods
2.1 Data collection
The primary steps of the suggested financial classification approach were data collection, analysis, processing, and evaluation. Physical, roadway, and traffic damage were among the identified classes used to categorize the damage. Ultimately, the results of this study are evaluated and interpreted, and are shown in the Results section.
The physical financial impacts were collected by ODOT during its Oregon Wildfire Response and Recovery (OWRR) efforts during and after the wildfires in 2020. The cost impacts were captured, detailed, and summarized on the Federal Highway Administration (FHWA) DDIRs that ODOT completed along each highway, (Appendices A-F in S1 Text). Each cost impact is itemized, with additional detail on site location as well as a description of the damage. The categories of financial impacts were established through reviewing, and summarizing the monetary data presented in the ODOT provided DDIRs. The criteria utilized during the categorization included examining and interpreting the damage data identified and combining similar damage groups to reduce and simplify the number of overall classifications. The data received from ODOT for this study were in standard units and preserved as such in the appendices but are later converted to metric units. Traffic count information [38] was used to acquire milepost and ADT count data, which was then extrapolated to correspond with the various closure zones on each of the highways. To provide accurate ADT statistics, along with estimated dates of each wildfire’s onset and containment and hectares burned for each fire, it was necessary to identify the areas of road closure and align them with the relevant ADT milepost information.
2.2 Data analysis
Physical financial parameters for this assessment includes defining the length of highway impacted as well as the average daily traffic (ADT) counts affected by wildfires and their estimated costs. Including the ADT counts as part of the financial consequences of the wildfires captures the increased km and time required to reach the same destinations when using the shortest detour route to get there. Financial roadway impacts to the highways due to the wildfires include the costs of hazard tree removal, pavement damage and the subsequent repairs, slope/scaling damage from rocks, dirt, and debris that slid onto the road, and structural damage that required repair along the highways. The traffic control financial data include the estimated number of highway kilometers that were closed as well as the distance (km) of highway requiring traffic control during repair operations. While the cost of traffic impacts might normally be calculated in dollars, in evaluating the impacts of reduced ADT along these highways during the wildfires, the cost involves the increasing distance and time vs. an actual dollar value.
2.3 Data processing
The purpose of the proportional statistical calculations (Appendix G in S1 Text) is to provide a baseline for determining average costs (ACi) for repairing/maintaining highways in the wake of the wildfires based on the linear distance of highway impacted and the potential cost impacts related to the ADT on the highways. The cost of the roadway impacts was developed by dividing the temporary and permanent repair costs from each highway by the amount of roadway damage, while the damage cost per km of highway was obtained by similarly dividing the temporary and permanent repair costs from each highway by the distance (km) of highway impacted (Appendix G in S1 Text). The cost of traffic control per km of highway was calculated by dividing the temporary and permanent repair traffic control costs from each highway by the linear distance (km) of highway impacted (Appendix G in S1 Text). The distances (km) over which AADT cost impacts were accrued on each of the east-west highway routes were examined by first multiplying the AADT decrease (2015–2019)– 2020) by the increased distance in km. The minutes of AADT cost impacts of each of the east-west highway routes were examined by AADT decrease (2015–2019)– 2020, which was then multiplied by the increased number of minutes and then converted to hours. The reduced average annual daily traffic (RAADT) from 2015 to 2020 was calculated by averaging the previous 5-year (2015–2019) AADT data for each of the highways using the same MP range as previously noted in the highway distance affected. The 2020 AADT was then divided by this 5-year average to the RAADT percentage due to the wildfires (Appendix G in S1 Text).
A hypothesis test using the statistical programming language "R" is used to see if there is a relationship between the total cost of damage incurred on the highway—the outcome variable (the total length of the highway impacted (km)), which will be displayed in scatterplot matrix format, and the different damage factor costs of the predictor variables (hazard tree damage (km), pavement damage (km), slope/rock scaling damage (km), and structural damage (km). The question being examined is: do the financial damage elements in our research provide an indication of the total cost of kilometers that wildfires have affected on highways?
- Null Hypothesis–there is no relationship between the predictor variables and the outcome variable.
- Alternative Hypothesis–the null is not true.
3. Results
Physical financial impacts
The physical components of the highway’s financial damage are identified in Fig 2, which are broken down by the length (km) of highway impacted by each of the elements. While the length of highway impacted by structural damage appears to be the most significant, surprisingly, the length of highway impacted by actual pavement damage is showing the least amount of damage. The highway financial damage, while quantified for each individual highway within each wildfire were analyzed, their location, whether it was in the middle, or the edge of the wildfire wasn’t a definitive factor in determining greater or less damage depending on the size or location of the highway within the wildfire. The location of each of the highways within the state of Oregon–the northern, central, or southern portion don’t seem to have any direct pattern concerning the amount of damage sustained per km of highway. Another item of note is that the size of the wildfire (ha) doesn’t appear to have any bearing on the km of highway impacted by the various sections of damage (Appendix H1 in S1 Text).
To provide a more thorough grasp of the study problem, a mixed-methods methodology approach was used combining both qualitative and quantitative research procedures. Initially we used a quantitative approach to quantify the highway financial damage through a proportional statistical breakdown—equations, followed by a linear regression analysis to determine if any relationships actually existed between the different types of damage and the length of highway affected. This was followed up with a qualitative review through discussions/interviews with ODOT emergency personnel.
The length of highway closed (km) in relation to the overall length of highway impacted (Fig 3) shows a high of 100%, a low of 21%, and an average of 62%. No clear trend in the length of affected or closed route (km) for any of the highways located in the northern, middle, or southern regions of the state is evident. It is also noteworthy that the extent of the wildfire–(area burned (ha)) seems to be unrelated to the length (km) of roadway affected or closed because of the highway damage (Appendix I in S1 Text).
Fig 4 provides the estimated emergency repair costs in relation to the length of highway (km) closed during the wildfires, by highway. Over 307 km of these highways closed due to the wildfires, and the average emergency repair cost was $94,531 per km (Appendix J1 in S1 Text). The length (km) of highway closed has no apparent connection to the amount of financial damage sustained on each highway.
The financial cost data were put through the hypothesis test to determine if there is a relationship between the predictor variables and the outcome or response variable (Fig 5). Will the predictor variables (hazard tree removal costs (HTC), pavement cost (PC), slope/rock scaling costs (SRSC), and structural costs (SC)) be related to the outcome variable, total cost of repairs (T1)? The next step is determining which predictor variables will make it to the final model:
> Plot (c (T1, HTC, PC, SRSC, SC), fit = "lm")
The scatterplot matrix (Fig 6) shows how each predictor variable (HTC, PC, SRSC, and SC) is related to the outcome variable (T1) and each other. It appears that HTC has a positive correlation of 1 to T1 as well as SRSC, which also has a positive correlation of 1. PC has a very small positive correlation to T1 at 0.03 while SC has a negative correlation at -0.53. (It is interesting to note that the HTC and SRSC appear to have a strong correlation to each other at 0.99). While HTC and SRSC have positive relationships towards T1, their correlation does not prove causation. Here are the two predictor variables that will appear in the final model:
> Plot (c(T1, HTC, SRSC), fit = "lm")
Next, the regression model (Table 2) is evaluated to determine if there is a significant relationship between T1 and HTC and SRSC.
> reg(T1~HTC)
One of the most significant pieces of information we can analyze from the regression analysis is the p-value. The p-value assumes the null hypothesis (that there is no relationship between variables) is true and is the probability that an observed difference between variables might have occurred by random chance. This can indicate to us whether there is a statistically significant relationship between predictor and outcome variables.
- For HTC, the p-value is 0.00.
- For SRSC, the p-value is 0.00.
To analyze the p-value, an adoption criterion of 5% probability (as it relates to the normality curve assuming a 95% confidence interval) which is the alpha probability that defines an unlikely even has occurred given our assumption that the null hypothesis is true.
HTC
- p-value = 0.00 < a = 0.05
SRSC
- p-value = 0.00 < a = 0.05
In both cases, assuming the null hypothesis, that there is no relationship between the predictor variables and outcome variable, an unlikely event occurred, so the null hypothesis is rejected (HTC and SRSC do not have a relationship with T1) and find that results of the test support the alternative hypothesis. Based on this test, it is feasible that the HTC and SRSCs have a statistically significant relationship to T1.
The second test was to determine if the predictor variables (hazard tree damage costs (HTC), pavement damage costs (PC), slope/rock scaling damage costs (SRSC), and structural damage costs (SC) are related to the outcome variable (Size of fire) as shown in Fig 7.
We need to evaluate which predictor variables will make it to the final model:
> Plot(c(Size (ha), HTC, PD, SRSD, SD), fit = "lm")
The scatterplot matrix shows how each predictor variable (HTC, PC, SRSC, and SC) is related to the outcome variable (Size) and each other. It appears that PC has a positive correlation to Size (0.27) as well as SC, which has a positive correlation of 0.42. HTC has a negative correlation to Size at -0.42 and SRSC has a negative correlation at -0.44. (It is interesting that HTC and SRSC appear to have a strong correlation to each other at 0.99).
While PC and SC have positive relationships towards T1, their correlation does not prove causation. Here are the two predictor variables shown in Fig 8 display how they will appear in the final model:
> Plot(c(Size, PC, SC), fit = "lm")
Next, the regression model shown in Table 3 will be evaluated to determine if there is a significant relationship between T1 and PC and SC.
> reg(Size~PC)
> reg(Size~SC)
The p-value assumes the null hypothesis (that there is no relationship between variables) is true and is the probability that an observed difference between variables might have occurred by random chance. This can indicate to us whether there is a statistically significant relationship between predictor and outcome variables.
- For PC, the p-value is 0.658.
- For SC, the p-value is 0.481.
To analyze the p-value, we adopt a criterion of 5% probability (as it relates to the normality curve assuming a 95% confidence interval) which is the alpha probability that defines an unlikely even has occurred given our assumption that the null hypothesis is true.
PC
- p-value = 0.658 > a = 0.05
SC
- p-value = 0.481 > a = 0.05
In both cases, assuming the null hypothesis, that there is no relationship between the predictor variables and outcome variable, a likely event occurred, so we fail to reject the null hypothesis that PC and SC do not have a relationship with the Size of the wildfire.
ADT impacts from the wildfires to the highways are shown in Figs 9 and 10, which caused an average increase of travel distance of 24 km and 17 minutes in travel time on alternative highway detour routes (Appendix K in S1 Text). This equated to an estimated average increase of 11% in the highway distance traveled on the detours as well as an estimated average increase of 11% in the time needed to reach the same location on the 307 km of highway that were closed (Appendix K in S1 Text). One item that is noticeable in this comparison is that except for Highway 224, all the detour routes are longer in length and require more time to travel. Highway 224’s detour route is the same distance as its existing route and the time it takes on the detour route is less than using the existing route. Even though the highways were closed for prolonged periods of time, the additional distance and time required seems relatively minor considering the overall ADT of these highways, who’s routes are essential for access to Oregon’s transportation system in terms of business, freight, and tourism.
3.2 Roadway financial impacts
Fig 11 presents the estimated expense for emergency repair hazard tree removal, compared to the length of highway (km) impacted by hazard trees due to the wildfires, by highway. The average estimated emergency hazard tree removal cost along these routes was $1,666,667 per km (Appendix L1 in S1 Text). The estimated amount of money needed to remove hazards trees from each highway does not seem to be related to the length (km) of the affected highway (Appendix A in S1 Text).
The length of the highway (km) impacted by pavement damage due to the wildfires is plotted against the expected cost of pavement damage emergency repairs in Fig 12. Along these roads, the average estimated cost of emergency repair pavement damage was $8,243 per km (Appendix M1 in S1 Text). The anticipated cost of repairing the pavement damage on each roadway is not related to the damaged highway’s length (kilometers) (Appendix B in S1 Text).
Fig 13 shows the length of highway (km) impacted by slope/rock scaling damage due to the wildfires against the projected cost of slope/rock scaling damage emergency repairs. The average projected cost of emergency repairs for rock scaling and slope degradation along these roads was $35,203 per km (Appendix N1 in S1 Text). Once more, the estimated cost of the rock scaling/slope damage for each road is independent of the length (km) of the affected highway (Appendix C in S1 Text).
Fig 14 shows the length of the highway (km) impacted by structural damage due to the wildfires against the estimated cost of structural damage emergency repairs (Appendix O1 in S1 Text). It shows that the average estimated cost of structural emergency repairs along these roads was $8,660 per km. Once more, the length (km) of the impacted highway apparently has no bearing on the estimated cost of the structural damage to each route. In addition, structural damage occurred to Bad Banks Bridge (State Bridge #06761) along OR22, MP 36.74, which included structural damage to the concrete inlet, wing walls, and apron, culvert liner, and retaining wall; galvanized coating melted off of manhole cover; melted HDPE diversion pipes; and burned trees that pose a danger to the bridge (Appendix D in S1 Text).
3.3 Traffic financial impacts
The highway length (km) affected by the wildfires is plotted against the projected cost of emergency traffic management demands in Fig 15. The average estimated cost of traffic control along these highways was $9,657 per km, while the total estimated cost of emergency and permanent traffic control was $3,172,310.as shown in Appendix P1 in S1 Text. No apparent correlation exists between the financial cost of traffic control on each roadway and the length (km) of the highway that needs it (Appendix E in S1 Text).
Corridor status maps show the work being done on each of the highways, the regular speed limit vs. the restricted speed limit, and the date, time, and MP span of the closure. Each of the highways utilized road closure status maps that were updated during the 2020 wildfires (Appendix F in S1 Text). Road closures fell into four categories:
- Full closure
- Critical services that allow emergency and service vehicles only
- Partial access
- Traffic restored with some ongoing repairs.
Fig 16 shows the AADT for the 2015–2019 period vs. that for 2020. The AADT dropped by a total of 125,197 vehicles for 2020 (Appendix Q1 in S1 Text). Due to each highway being closed for various periods of time during the wildfires, a drop in AADT was expected. Fig 17 expresses AADT for each highway into percentages for quick comparison and averaged 14% overall for all of the highways (Appendix Q1 in S1 Text). The wildfires identified in this study averaged an additional 2,979,698 km of highway detour routes and 35,055 hours of time required using these detour routes for the highways being evaluated. These additional km and hours of time are identified as the worst-case scenario, using all the AADT counts that have been determined (Appendix Q1 in S1 Text).
The use of the alternative detour routes will not only add distance and time to the AADT, but also increase fuel costs due to the additional distance involved. The average price of gasoline in Oregon in 2020 was $2.17 per gallon or $0.57 per liter [39]. Using this price as a baseline, and an average of 10.92 km per liter [40] for fuel efficiency, the additional gasoline cost due to the alternative detour routes can be established. This averaged to an approximate increase of $155,534 in fuel costs along with an additional average of 2,979,698 km due to the use of alternative detour routes.
4. Discussion
This study is a first attempt to evaluate the financial risks, dangers, and harm to Oregon’s road infrastructure caused by megafires, which are frequently disregarded when calculating hazard risk, vulnerability, and resilience. The economic effects of wildland fires, which include losses of residential and commercial property (homes, personal items, commercial structures, and inventories), are frequently discussed in archival works that examine how climate change influences infrastructure damage. Monitoring the cost of wildfires can be used to evaluate the return on investment for fighting them, considering both direct and indirect losses that include lost wages, consumer spending, and future infrastructure investments, Appendix R in S1 Text. Unfortunately, there is a dearth of research on the financial damage to highway infrastructure due to wildfires.
With the growing number of devastating megafires in Oregon and the subsequent destruction and financial impacts caused to highways, the data generated in this work makes it possible to identify specific areas of damage to highways with a means of rapidly determining cost impacts for estimating overall financial damage ramifications to local agencies, state governments and more specifically federal agencies such as FHWA and FEMA.
Since previous data on highway damage from wildfires is extremely limited, the results in this study were compared to the only other two megafires in Oregon that occurred in 2015 (Appendix H2 in S1 Text) that collected similar data for reimbursement from the federal government that are summarized in Table 4.
To achieve the research goals, three financial features were analyzed in the study: impacts to physical characteristics, roadways, and traffic. Each of these three features included sub-features that included the costs of the distinct types of damage. The financial impacts quantified in the 2020 Labor Day event came from the ODOT DDIRs and included a) impacts to physical characteristics, b) impacts to roadways and c) impacts to traffic. Factors identified in each of these 3 categories include: a) physical characteristics—total overall estimated cost (temporary and permanent), average (temporary and permanent) repair cost per km of highway affected, and the increased ADT expenses in both the distance and the amount of time required while using detours; b) roadways—total estimated cost (temporary and permanent) of hazard tree removal, pavement damage, slopes/rock scaling damage and structure damage; c) traffic—total estimated (temporary and permanent) cost of traffic control, and the increased ADT average km of highway detour routes and hours of time required using these detour routes. Multiple proportional statistics were established (Appendix G in S1 Text), which provide a baseline for cost impacts for each of the identified characteristics. The quantification of the financial impacts for these features showed limited relationships between the size of the wildfires, the subsequent features, and the financial damage. The major results from this research can be summarized as follows:
- ○ Physical Highway Costs (Appendix J1 in S1 Text; Appendix K in S1 Text)
- The total estimated cost of emergency and permanent repairs due to the wildfires was $44,894,471.
- The average emergency repair cost per km of highway affected was $94,531, while the average permanent repair cost per km of highway affected was $51,705.
- As a result of the wildfires, ADT expenses increased by an average of 11% in both the distance and the amount of time required while using the detours.
The physical cost impact results showed millions of dollars were spent on emergency repairs on each of the four individual highways damaged in 2020, which were each closed for several weeks or in some cases months, whereas the two highways damaged in the wildfires that occurred in 2015 (Appendix J2 in S1 Text) had significantly lower costs but didn’t close either road during the wildfires. While the fires in 2020 and in 2015 all had traffic control required during the wildfires, it’s obvious that closing the roads is an indication of more financial damage to the roadways vs. not having to close the roads. Without any roads being closed during the two 2015 megafires, any comparisons to the average repair costs per km either temporary or permanent against the costs determined during the 2020 megafires would be purely hypothetical without any real basis. Since the roads weren’t closed during the 2015 wildfires there were no detour routes required, therefore there was no additional distance or time needed utilizing detours and again, there isn’t a comparison to detours required during the 2020 wildfires. These results reflect the idea that the length of highway impacted, closed, requiring traffic control or the size of the wildfire are not related to the amount of financial damage done to the highway.
- Roadway Impacts Costs (Appendix L1 in S1 Text; Appendix M1 in S1 Text; Appendix N1 in S1 Text; Appendix O1 in S1 Text)
- The total estimated cost of emergency and permanent hazard tree removal costs was $5,000,000.
- The total estimated cost of emergency and permanent pavement damage was $428,660.
- The total estimated cost of emergency and permanent slopes/rock scaling damage was $2,112,200.
- The total estimated cost of emergency and permanent structure damage was $6,216,640.
Roadway cost impact results provided limited information that could be analyzed due to the amount of missing quantity estimates of the damage. With the exception of the Beachie Creek wildfire on highway 022 in 2020, all of the other wildfires, both in 2020 and 2015 (Appendix L2 in S1 Text, Appendix M2 in S1 Text, Appendix N2 in S1 Text, Appendix O2 in S1 Text) were missing pertinent data in one form or another along the highways, which made comparisons to one another difficult and caused a couple of different issues. This lack of consistently complete information on the DDIRs makes the comparison and analysis somewhat capricious primarily because a) the amount of data available is already modest to begin with and b) using only partial sets of complete data doesn’t adequately capture the overall dependability of the financial cost impacts. Instead of being able to develop equations based upon damage to all four highways, each hazard was evaluated on anywhere between one to four highways depending on if the highway DDIR’s had captured the financial damage impacts. The other issue is that it was impossible to compare the data to the 2015 wildfires because there were no estimated quantities of damage captured during those wildfires, just a generalization of the damage. Even so, the integration of these proportional statistics, while based upon limited available figures, should be built into decision making and can be adopted for future financial damage assessment to highways from wildfires.
- Traffic Impact Costs (Appendix P1 in S1 Text; Appendix Q1 in S1 Text)
- The total estimated cost of emergency and permanent traffic control was $3,172,310.
- The ADT averaged an additional 2,979,698 km of highway detour routes and 35,055 hours of time required using these detour routes for the highways being evaluated.
The traffic cost impact results indicated no relation between the length of highway impacted and the amount of traffic control required on the roadway. Again, only three of the four highways in 2020 provided data concerning traffic control costs, while only one of the two wildfires in 2015 (Appendix P2 in S1 Text) had traffic control data. Comparison of such a small amount of information is challenging and would benefit from further additional research and while the proportional statistics are based upon very limited data, it is feasible to interpret costs using the equations developed here. Perhaps the most interesting part of the AADT results was the comparison between the 2020 and 2015 wildfires. During the 2020 wildfires, all of the highways experienced a reduction in the AADT due to the wildfires, but the highways impacted by the 2015 wildfires actually increased their AADT (Appendix Q2 in S1 Text). The obvious difference between the two data sets is that the 2020 wildfires closed highways for various lengths of time diverting traffic to other roadways, while the 2015 wildfires didn’t close the highways. Another reason that the AADT may have increased during 2015 is the allure of onlookers getting a chance to see the wildfire firsthand during and after they were contained. Limitations of the 5-year Average Annual Daily Traffic (AADT) include a) small number of years for assessment comparison, b) potentially conceal daily fluctuations in demand, seasonal variations, and special events and c) challenging to capture any micro-trends on the various highways. Further studies can utilize the analysis provided here to capture and incorporate the AADT impacts to not only wildfire events, but other natural disasters that impact highways as well.
Both direct and indirect expenditures come into play when trying to capture the costs associated with wildfire damage–specifically to roadways. Direct expenditures can include costs for repairing and cleaning roadways that are primarily linked to the physical effects of wildfire occurrences. Indirect costs, on the other hand, are the additional economic effects of a wildfire event that go beyond the physical destruction and include losses and expenses such as the loss of use of transportation infrastructure, which can hinder transit, power, water, health, and public safety, as well as emergency response costs, engineering evaluations, and environmental impacts. The extra time and distance needed to reach the destination on a detour are examples of the indirect costs of a road closure.
Despite the practical benefits of this study, several limitations need to be addressed in future research. While more than 20 wildfires burned over 1,000 acres or 405 ha in Oregon in 2020, this study evaluated only the four highways impacted by the five megafires in Oregon during the 2020 Labor Day wildfires. Supplementary research should include the analysis of more highways affected by megafires within a geographical region.
Another limitation was the location of the highways in relation to the size (hectares burned) of the wildfires. Fig 1 shows that two of the highways—224 and 022 were located toward the edge of the wildfire, while the other two highways—126 and 138 were situated more centralized within the wildfire footprint. More research to determine how much financial damage occurs to highways depending on the highway location relative to the wildfire core could be valuable. Information from the DDIRs was missing for each of the distinctive features analyzed in the study (physical characteristics, roadway impacts, and traffic impacts) along with their sub features. Future financial cost research would benefit from more complete information in each of these categories. Additional limitations include the ADT information gathered for this report. While this report analyzed the 5-year ADT information prior to the 2020 Labor Day wildfires, the more years that are evaluated, the greater the accuracy of the reduction in ADT in relation to wildfire impacts can be calculated. Future research would benefit from more comprehensive information in each of the categories.
Some of the challenges that this research faced included (a) initially obtaining the research data, (b) identifying and breaking down the data into individual features/sub-features, and (c) recognizing that much of the financial damage information provided for each wildfire was consistently incomplete.
Without a doubt, the most challenging issue to solve has been getting the scant data needed for this study. The initial data from Oregon for roadway damage brought on by the megafires were extremely difficult to obtain. Since there was no federal, state, or regional database to begin with, the first challenge was figuring out what kind of roadway damage data to request. Finding the locations of the roadway damage records from the megafires was still necessary once the damage categories had been established. We were provided a few DDIRs from ODOT’s public information office, however, we were then required to request additional information using their public records information system.
In addition to the requests for public information records, Oregon emergency personnel were contacted to provide further clarification through phone conversations and emails to obtain the relevant damage information. Prior to this study, no one had ever asked for or requested this type of information. Upon examining Highway 22 in Oregon, which was affected by the Beachie Creek and Lionshead megafires, we discovered that ODOT maintenance staff did not assess the damage along the highway according to each wildfire, which makes sense. However, this meant that we had to ascertain the locations of the beginning and end of each wildfire, correlate those locations with mileposts along the highway, and break out the damages based on the wildfire they were within.
The requirement to keep describing the wildfire damage risk to highways in greater depth is another problem brought about by our effort. The study has the benefit of tackling a real-world problem that hasn’t gotten enough attention thus far: wildfires and the harm they do to the infrastructure of roads. Every year, we deal with this practical issue on a regular basis. This research can be used to understand how this topic influences average daily traffic (ADT) on the highways and the additional costs associated with it, in addition to assessing damage to the roads.
Future work should entail surveying and interviewing emergency personnel in different states or regions, as well as obtaining information from public records domains, including upfront initial costs and subsequent reimbursement costs from the federal government, and compiling such data in a standardized database. Regardless, the upcoming improvements to highway financial damage assessment and foundation analysis will improve our ability to anticipate and lessen the risk of wildfire damage to roadways.
Transportation authorities must comprehend the financial aspects of the damage to the transportation network to plan effectively for and manage emergencies. To have a fair understanding of the true impact of huge wildfires, which frequently exceeds the standard impact indicators, studies of the economic effects of catastrophic wildfires must be thorough and encompass all economic aspects. The purpose of our work aims to provide emergency planners at agencies like ODOT and other governance emergency management organizations with a clear and concise set of information regarding the highway financial damage caused by megafires during the 2020 Labor Day wildfires, regardless of their background in transportation. The results can theoretically support highway damage identification and evaluation as well as future infrastructure investments.
5. Conclusions
This paper assembled and reviewed the available data needed for the analysis of megafire highway financial risks that occurred in Oregon during the 2020 Labor Day wildfires. The findings show a weak or non-existent correlation between the extent of the wildfire’s damage and the highway’s financial costs, but there is enough proof to support professionals’ and practitioners’ assessments and predictions about the financial effects megafires will have on roadways. In general, none of the highways affected by the megafires clearly stand out as having the most damage per kilometer of highway, nor do they exhibit any kind of pattern of greater damage or traffic impacts based on the highway’s location within the state, the length of the highway that was affected or closed, or the location of the highway within the wildfire.
As this work illustrates, the cost impacts associated with damage to highways due to megafires in Oregon, a U.S. state with an average infrastructure, a sparse population, and a potentially growing susceptibility to extended dry spells, can be substantial. The specific conclusions drawn from the analysis in this paper are:
- Public health, evacuation/mobility, suppression/fire management, and property loss during the event are the main areas of focus for current research on the cost implications of wildfires related to transportation; direct financial repercussions are given very little attention.
- The length of the route and the intensity of the wildfire are two criteria that can affect how financially vulnerable transportation infrastructure is to megafires. Other aspects include the physical characteristics of the roadway and traffic impacts.
- New techniques for assessing financial threats to highways using historical data and new technologies that can assist agencies in rapidly grasping quantity and damage estimates will become indispensable as wildfires continue to overwhelm the transportation system.
The main limitations with this study’s methodology and data analysis include the following: a) limited number of impacted highways evaluated, b) only focusing on megafires, c) location of the highways in relation to the magnitude of the wildfires, d) incomplete DDIR information, e) limiting the AADT review to 5-years. These factors on the whole influenced the conclusions through a) the utilization of a very small data set, b) missing potentially vital highway damage information by only focusing on megafires, c) determining if further research is needed to ascertain if the location of the highway within a wildfire has any impact on the amount of highway damage incurred, d) complete DDIR information would strengthen and solidify the results that have been established, e) increasing the AADT years of analysis could potentially alter the estimated financial impacts identified
Future assessments of the monetary damage that wildfires cause to highways, which are now constrained by a lack of data, are needed. The computations of developed proportional statistics should be extended to more case studies, as should the standardization of financial estimates of the costs for road characteristics for use in estimating the risk of future wildfires. Consideration must be given to methods of projecting those estimates with the ratings of active wildfires, to broaden and enhance this initial pilot study.
Supporting information
S1 Text. Appendices.
Appendix A Hazard tree removal from the 2020 Oregon Labor Day wildfires. Appendix B Costs of pavement damage repairs from the 2020 Oregon Labor Day wildfires. Appendix C Slope/rock scaling damage from the 2020 Oregon Labor Day wildfires. Appendix D Structure damage from the 2020 Oregon Labor Day wildfires. Appendix E Costs of traffic control from the 2020 Oregon Labor Day wildfires. Appendix F Road closures caused by the 2020 Oregon Labor Day wildfires. Appendix G Highway Financial Impact Equations. Appendix H1 Physical Components of 2020 Labor Day wildfires in Oregon. Appendix H2 Physical Components of 2015 wildfires in Oregon. Appendix I Length of Highways Impacted/Closed during the 2020 Labor Day wildfires in Oregon. Appendix JI Length of Highways Closed & Emergency Repairs on 2020 Labor Day wildfires in Oregon. Appendix J2 Length of Highways Closed & Emergency Repairs on 2015 wildfires in Oregon. Appendix K Highway ADT 2020 Cost Impacts. Appendix L1 KM of Highway Impacted vs. Estimated Cost of Hazard Tree Removal in 2020 Labor Day wildfires in Oregon. Appendix L2 KM of Highway Impacted vs. Estimated Cost of Hazard Tree Removal in 2015 wildfires in Oregon. Appendix M1 KM of Highway Impacted vs. Estimated Emergency Repair Cost of Pavement Damage in 2020 Labor Day wildfires in Oregon. Appendix M2 KM of Highway Impacted vs. Estimated Emergency Repair Cost of Pavement Damage in 2015 wildfires in Oregon. Appendix N1 Slope/Rock Scaling Damage in 2020 Labor Day wildfires in Oregon. Appendix N2 Slope/Rock Scaling Damage in 2015 wildfires in Oregon. Appendix O1 Structural Damage in 2020 Labor Day wildfires in Oregon. Appendix O2 Structural Damage in 2015 wildfires in Oregon. Appendix P1 KM of Highway Impacted vs. Estimated Emergency Traffic Control Cost in 2020 Labor Day wildfires in Oregon. Appendix P2 KM of Highway Impacted vs. Estimated Emergency Traffic Control Cost in 2015 wildfires in Oregon. Appendix Q1 AADT 2015–2019 vs. AADT 2020 in Oregon. Appendix Q2 AADT 2010–2014 vs. AADT 2015 in Oregon. Appendix R Economic Burden of Wildfire used to assess the Return on Investment into Wildfire Interventions.
https://doi.org/10.1371/journal.pclm.0000489.s001
(DOCX)
References
- 1.
Bayham J, Yoder JK, Champ PA, Calkin DE. The Economics of Wildfire in the United States [Internet]. Annual Reviews; 2022. https://doi.org/10.1146/annurev-resource-111920-014804
- 2.
Calkin DE, Thompson MP, Finney MA. Negative consequences of positive feedback in US wildfire management—forest ecosystems. SpringerLink; 2015. https://doi.org/10.1186/s40663-015-0033-8
- 3. Walls M, Wibbenmeyer M. How local are the local economic impacts of wildfires?. 2023. Available from: https://media.rff.org/documents/WP_23-03.pdf.
- 4. Thomas D, Butry D, Gilbert S, Webb D, Fung J. The costs and losses of wildfires. NIST Special Publication. 2017;1215(11).
- 5. Mass CF, Ovens D, Conrick R, Saltenberger J. The September 2020 wildfires over the Pacific Northwest. Weather and Forecasting. 2021; 36(5), 1843–1865.
- 6. Bertoletti AZ, Phan T, Campos do Prado J. Wildfire smoke, air quality, and renewable energy—examining the impacts of the 2020 wildfire season in Washington state. Sustainability. 2022; 14(15), 9037. https://www.mdpi.com/20711050/14/15/9037.
- 7. Juliano TW, Jiménez PA, Kosović B, Eidhammer T, Thompson G, Berg LK. Smoke from 2020 United States wildfires responsible for substantial solar energy forecast errors. Environmental Research Letters. 2022; 17(3), 034010.
- 8. Hoover K, & Hanson LA.mWildfire statistics. Congressional Research Service. 2021. Available from: https://apps.dtic.mil/sti/pdfs/AD1143321.pdf.
- 9.
Morton DC, Roessing ME, Camp AE, Tyrrell ML. Assessing the environmental, social, and economic impacts of wildfire. GISF Research Paper, 1. Forest Health Initiative. 2003. Available from: https://yff.yale.edu/sites/default/files/files/wildfire_report(1).pdf.
- 10. Orwig K. Strategies for funding wildfire mitigation. Risk Management. 2016; 63(5), 14–15. Available from: https://www.proquest.com/openview/fc86e37b27d8d7597f197ad0e80b51ef/1?pqorigsite=gscholar&cbl=47271.
- 11.
Zybach B, Brenner G, Dubrasich M, Marker J. U.S. wildfire cost-plus-loss economics project: The “one- pager” checklist. Advances in Fire Practices. Wildland Fire Lessons Learned Center. 2009. Available from: https://www.researchgate.net/profile/BobZybach/publication/237396187_US_Wildfire_Cost-PlusLoss_Economics_Project_The_One-Pager_Checklist/links/575ade8508ae9a9c955190b6/US-Wildfire-Cost-Plus-LossEconomics-Project-The-One-Pager-Checklist.pdf.
- 12.
Chester MV, Li R. Vulnerability of California roadways to post-wildfire debris flow. UC Office of the President: University of California Institute of Transportation Studies. 2020. Available from: https://escholarship.org/uc/item/60d0k700.
- 13. Chicco JM, Mandrone G, Vacha D. Effects of wildfire on soils: Field studies and modelling on induced underground temperature variations. Frontiers in Earth Science. 2023; 11.
- 14. Thompson MP, Calkin DE, Finney MA, Gebert KM, Hand MS. A risk-based approach to wildland fire budgetary planning. Forest Science, 2013a; 59(1), 6377.
- 15. Mouratidis A. Road adaptation to climate hazards: Guidelines for cost-effective measures. Journal of Earth and Environmental Sciences Research. 2020; 2(3), 1–6. Available from: https://www.onlinescientificresearch.com/articles/road-adaptation-to-climate-hazardsguidelines-for-costeffective-measures.pdf.
- 16.
Hoover K. Federal assistance for wildfire response and recovery. Congressional Research Service. 2015. Available from: https://union-county.org/cwpp/Project%20File/Reference_materials/Fire_Policies_Budget_info/FEMA /R41858.pdf.
- 17. Boyd W. Climate liability for wildfire emissions from federal forests. Ecology Law Quarterly. 2021;48(4), 981–1014.
- 18. Mostafiz RB, Friedland CJ, Rohli RV, Bushra N. Estimating future residential property risk associated with wildfires in Louisiana, USA. Climate. 2022;10(4), 49.
- 19.
Liu J. Wildfire: Review of life, property, and economic damages. Donald C. Hellmann Task Force Program, Henry M. Jackson School of International Studies, University of Washington. 2021;38–47. Available from:https://jsis.washington.edu/wordpress/wpcontent/uploads/2021/03/21_JSIS-495H_Final-Report_Montgomery1.pdf#page=39.
- 20. Richardson LA, Champ PA, Loomis JB. The hidden cost of wildfires: Economic Valuation of health effects of wildfire smoke exposure in Southern California. Journal of Forest Economics. 2012;18(1), 14–35.
- 21. Chen H, Samet JM, Bromberg PA, Tong H. Cardiovascular health impacts of wildfire smoke exposure. Particle and Fibre Toxicology. 2021;18(1). pmid:33413506
- 22. Rappold AG, Reyes J, Pouliot G, Cascio WE, Diaz-Sanchez D. Community vulnerability to health impacts of wildland fire smoke exposure. Environmental Science & Technology. 2017;51(12), 6674–6682. pmid:28493694
- 23. Miller N, Molitor D, Zou E. Blowing smoke: Health impacts of wildfire plume dynamics. 2019. Available from: https://giesbusiness.illinois.edu/nmiller/documents/research/smoke.pdf.
- 24. Kim MK, Zhu E, Harris TR, Alevy JE. An LP-SAM approach for examining regional economic impacts: An application to wildfire disasters in southeast Oregon. Review of Regional Studies. 2012;42(3), 207–221. Available from: https://digitalcommons.usu.edu/appecon_facpub/1254/.
- 25.
Maher AT. The Economic Impacts of Sagebrush Steppe Wildfires on an Eastern Oregon Ranch. Masters Thesis, Oregon State University. 2007. Available from: https://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/7s75dg19x?loc ale=en.
- 26. Thompson MP, Vaillant NM, Haas JR, Gebert KM, Stockmann KD. Quantifying the potential impacts of fuel treatments on wildfire suppression costs. Journal of Forestry. 2013b;111(1), 49–58.
- 27. Zhai J, Kuusela OP. Estimating price dynamics in the aftermath of forest disturbances: The Biscuit fire in southwest Oregon. Forest Science. 2020; 66(5), 556–567.
- 28.
Gorte R. The rising cost of wildfire protection. Bozeman, MT, USA: Headwaters Economics. 2013. Available from: https://www.baileyhealthyforests.org/wp-content/uploads/2013/12/fire-costsbackground-report.pdf.
- 29. Sánchez JJ, Holmes TP, Loomis J, González-Cabán A. Homeowners willingness to pay to reduce wildfire risk in wildland urban interface areas: Implications for targeting financial incentives. International Journal of Disaster Risk Reduction. 2022;68.
- 30.
Nielsen-Pincus M, Ellison A, Moseley C. The effect of large wildfires on local labor markets. Scholars’ Bank. University of Oregon. 2012. Available from: https://scholarsbank.uoregon.edu/xmlui/handle/1794/19384.
- 31. Davis EJ, Moseley C, Nielsen-Pincus M, Jakes PJ. The community economic impacts of large wildfires: A case study from Trinity County, California. Society & Natural Resources. 2014;27(9), 983–993.
- 32.
Roper D. Understanding Wildfire Mitigation Behavior of Central Oregon Homeowners. Masters Thesis. Oregon State University. 2015. Available from: https://ir.library.oregonstate.edu/concern/graduate_projects/47429b857.
- 33. Mietkiewicz N, Balch JK, Schoennagel T, Leyk S, St. Denis LA, Bradley BA. In the line of fire: Consequences of human-ignited wildfires to homes in the US (1992–2015). Fire. 2020;3(3), 50.
- 34.
Colburn L, Marsters L, Gartner T. 3 lessons on wildfire resilience and recovery from Oregon’s Santiam Basin. World Resources Institute. 2022. Available from: https://www.wri.org/update/3-lessons-wildfire-resilience-and-recovery-oregons-santiambasin.
- 35.
Cohen J. The wildland-urban interface fire problem: A consequence of the fire exclusion paradigm. US Forest Service Research and Development. 2008. Available from: https://www.fs.usda.gov/research/treesearch/33787.
- 36. Oregon Natural Hazards Mitigation Plan. State of Oregon Promulgation. 2020. Available from: https://www.oregon.gov/lcd/nh/pages/mitigation-planning.aspx?utm_source=LCD&utm_medium=egov_redirect&utm_campaign=https%3A%2F%2Foregon.gov%2Flcd%2Fhaz%2Fpages%2Fnhmp.aspx.
- 37. USGS. USGS National Map Viewer. 2020. Available from: https://apps.nationalmap.gov/viewer/.
- 38. Oregon Department of Transportation. Traffic counting. 2023. Available from: https://www.oregon.gov/odot/data/pages/traffic-counting.aspx.
- 39. Dodds M. 2020 Oregon gas price news. AAA Oregon/Idaho. 2022. Available from: https://info.oregon.aaa.com/2020-oregon-gas-price-news/.
- 40.
Vehicle Technologies Office. FOTW# 1177, Preliminary data show average fuel economy of new light-duty vehicles reached a record high of 25.7 mpg in 2020. Office of Energy Efficiency & Renewable Energy. 2021. Available from: https://www.energy.gov/eere/vehicles/articles/fotw-1177-march-15-2021-preliminarydata-show-average-fuel-economy-new-.light#:~:text=Preliminary%20data%20for%20EPA’s%202020,light%2Dduty%20vehicle%20fuel%20econom.