Evaluating rural Pacific Northwest towns for wildfire evacuation vulnerability

Authors: Alex Dye*, Oregon State University, John B. Kim, USDA Forest Service Pacific Northwest Research Station, Karin L. Riley, USDA Forest Service Rocky Mountain Research Station, Andrew McEvoy, USDA Forest Service Pacific Northwest Research Station, Fang Fang, University of Illinois
Topics: Spatial Analysis & Modeling, Applied Geography, Coupled Human and Natural Systems
Keywords: Wildfire, evacuation, Oregon, Washington, rural, roads
Session Type: Virtual Paper
Day: 4/10/2021
Start / End Time: 8:00 AM / 9:15 AM
Room: Virtual 7
Presentation File: No File Uploaded

Wildfire is an annual threat for many rural communities in the Western United States. In many severe events, evacuation is one course of action to gain safety from an advancing wildfire. Since most evacuations occur in a personal vehicle along the surrounding road network, the quality of this network is a critical component of a community's wildfire vulnerability. In this paper, we leverage a high-resolution spatial dataset of wildfire burn probability and mean fireline intensity to conduct a regional-scale screening of wildfire evacuation vulnerability for 696 Oregon and Washington rural towns. We characterize each town’s surrounding road network to construct four simple road metrics: 1) the number of paved lanes leaving town that intersect a fixed-distance circular buffer; 2) the variety of lane directions available for egress; 3) the travel area that can be reached within a minimum distance while constrained only to movement along the paved road network; and 4) the connectivity of intersections. We then combine the road metrics with two metrics of fire exposure of the surrounding landscape through which evacuation will occur: 1) burn probability and 2) mean fireline intensity. By combining the road and fire metrics, we create an overall composite score for each town. Using maps and tabular data, we examine spatial patterns, advantages, and disadvantages of each metric. All metrics are poorly intercorrelated, indicating that using just one measure of egress vulnerability may not fully capture road network characteristics and that a multi-metric geographic analysis may better reveal egress limitations.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login