Civil Engineering ETDs

Publication Date

Spring 4-6-2017


Past suppression-based wildfire management practices have increased the frequency and intensity of wildfires. Advocates for the re-introduction of natural wildfire regimes must also prioritize wildfire damage protection, especially for vulnerable communities located near forests. Areas where urban and forest lands interdigitate are called the Wildland Urban Interfaces (WUIs). In the United States, the area of the WUIs is increasing, making more people vulnerable to wildfires. By responding to four research objectives, this dissertation proposed and tested an integrated framework for wildfire risk mitigation decision making at WUIs. Decision makers who could benefit from the results of this dissertation include WUI homeowners, community planners, insurance companies, and agencies that provide financial resources for managing wildfire.

The first objective investigated the complex relationship between wildfire and property values in a WUI community affected by a catastrophic wildfire event. The analysis focused on evaluating whether the damage from a previous wildfire, and the risk from a potential future wildfire are negatively capitalized in the housing market of a WUI community. A Hedonic Pricing Method (HPM) was applied on homes in Los Alamos County located in Northern New Mexico. Los Alamos is the home of a highly educated and high income community which experienced the Cerro Grande fire in 2000. Results showed that wildfire damage has a negative impact on the housing price, whereas future wildfire risk is a positive driver in the Los Alamos housing market. These findings support the wildfire mitigation paradox that states that WUI homeowners tend to underinvest for mitigating wildfire risk on their properties.

The second objective investigated the optimal investment required for mitigating the vulnerability of residential buildings to wildfire. The optimal retrofit plan for individual homes was estimated using an integer programming method. The evaluation function for this optimization is based on a multi-attribute vulnerability assessment system that yields a wildfire vulnerability rating for all properties in the study area. A feasible solution to this optimization problem is one that decreases the vulnerability rating of the house to an acceptable rating. Additional data included: (i) vulnerability assessment cards of the properties, (ii) building and site characteristics of the properties, and (iii) unit costs of implanting appropriate retrofit measure on each element of the property. These datasets were collected for 389 properties in Santa Fe County’s WUIs. Using an integer programing model, the total cost of reducing the vulnerability ratings from “high” and “very high” to “moderate” vulnerability level was estimated for each property. To account for uncertainties in the costs of implementing a specific retrofit measure, a Monte-Carlo sampler was used to generate 2,400 cost scenarios from cost probability distributions. Using a regression analysis on the property data, a cost function for vulnerability mitigation through retrofitting was derived. The cost function allows estimation of the retrofitting cost per area of the house and considering the initial vulnerability rating of the house.

The third objective was to investigate wildfire optimal mitigation investment schedules for homeowners. Two types of investments for mitigation were analyzed, namely self-insurance and market insurance. Self-insurance is represented financially as the amount homeowners spend to implement retrofit measures to reduce their property’s vulnerability to wildfires. Market insurance is the transfer of wildfire damage liability to a third party or insurance company. The investment decision of homeowners over a multi-year investment plan considering the effects of budget and market insurance policy constraints was formulated. The effectiveness of self-insurance improvements was modeled as a damage probability function. Using a mixed-integer programming model, the optimal annual investment for market and self-insurance was estimated. The case study in this chapter demonstrated the effect of various parameters on the investment schedule of honeowners. This case study considered the time value of money and insurance companies’ contingency policies and budget constraints. The results showed that in the absence of budget constraints and mandates on mitigation, the homeowner’s optimal choice would be to fully invest on insurance and to purchase the broadest wildfire hazard insurance coverage. When a minimum mitigating retrofit effort is required by insurance companies, homeowners would invest more at the beginning of the period and decrease their investment through time. In this case results showed that a homeowner would achieve a higher expected value of investment than a homeowner with whose investments increase through time.

In the fourth objective, an Agent Based Model (ABM) is proposed to account for heterogeneity in homeowners’ attributes and behaviors when confronting wildfire risk hazard. The success of the community to reduce wildfire risk was evaluated by aggregating the impact of each individual agent’s behavior. The investment behavior of each homeowner for a five-year planning period was retrieved from the optimization model proposed in the third objective. A neighborhood of six homeowners was used to test the proposed ABM. When a wildfire occurs, the wildfire may or may not damage the property. Therefore, the loss accrued by each homeowner was stochastically simulated for each year in the simulation. The probability of loss was formulated as a function of the initial vulnerability rating of the property and the homeowners’ cumulative investment on mitigation. The analyzed scenarios considered different types of homeowners (i.e. mitigating or non-mitigating). The spatial impact of neighboring properties on the loss potential of a homeowner was modeled using a conceptual fire spread model based on a Cellular Automata propagation model. Results suggest that (i) the location of the property in combination with (ii) the investment behavior of the homeowner influences the neighborhood’s aggregate loss to wildfire. Policy-makers can better mitigate aggregate loss to wildfire by prioritizing certain locations over others.


Wildfire, Community, Risk, Vulnerability, Mitigation, Insurance, homeowner, decision making

Document Type




Degree Name

Civil Engineering

Level of Degree


Department Name

Civil Engineering

First Committee Member (Chair)

Dr. Vanessa Valentin

Second Committee Member

Dr. Robert Berrens

Third Committee Member

Dr. Brady Horn

Fourth Committee Member

Dr. Mark Stone