Driving behaviour during wildfire evacuation
Rohaert, Arthur LU
(2025)
In Report / Department of Fire Safety Engineering, Lund Institute of Technology, Lund University
- Abstract
- Introduction: Wildfires pose escalating risks to communities at the wildland--urban interface, often necessitating extensive evacuations. This thesis explores how driving behaviour during wildfire evacuations differs from routine conditions and how these differences can be modelled to improve traffic simulations, thereby supporting more informed planning and response decisions by authorities.
Objectives: Three main research objectives are addressed: (1) characterising how macroscopic traffic dynamics differ between wildfire evacuations and routine traffic conditions, (2) assessing the impact of wildfire smoke on car-following behaviour, and (3) developing a framework for traffic simulations that can support emergency planning for... (More) - Introduction: Wildfires pose escalating risks to communities at the wildland--urban interface, often necessitating extensive evacuations. This thesis explores how driving behaviour during wildfire evacuations differs from routine conditions and how these differences can be modelled to improve traffic simulations, thereby supporting more informed planning and response decisions by authorities.
Objectives: Three main research objectives are addressed: (1) characterising how macroscopic traffic dynamics differ between wildfire evacuations and routine traffic conditions, (2) assessing the impact of wildfire smoke on car-following behaviour, and (3) developing a framework for traffic simulations that can support emergency planning for and response during wildfire evacuations.
Methods and outcomes: (1) A dedicated data analysis method was developed to compare traffic dynamics during routine and evacuation scenarios, using traffic detector data from recent wildfire events in California. The analysis revealed that drivers move more slowly and leave larger gaps between vehicles during evacuations. If unaccounted for in simulations, these behavioural shifts can lead to overly optimistic evacuation time estimates.
(2) A custom driving simulator and virtual reality environment were designed to assess how reduced visibility from wildfire smoke affects driver behaviour. Results indicate that, under reduced visibility, drivers reduce their speed when travelling alone, but do not consistently adjust their following distance in congested traffic. These findings inform a visibility-sensitive car-following model.
(3) Finally, a simulation framework was proposed, combining data from real wildfire events, virtual reality experiments and evacuation drills. This framework supports the evaluation of evacuation strategies and planning decisions under varying conditions. A case study applied to a community of more than a thousand households in Colorado demonstrated the framework’s utility in assessing traffic management interventions.
Conclusion: By capturing evacuation-specific driving behaviours and their impact on traffic, this thesis provides practical approaches to enhance the realism of evacuation models, which can, in turn, support more reliable planning and safer wildfire evacuations. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/49e777d7-f159-45b6-9a3c-ada8efad7fb4
- author
- Rohaert, Arthur
LU
- supervisor
-
- Enrico Ronchi LU
- Jonathan Wahlqvist LU
- Silvia Arias LU
- opponent
-
- Prof. Cova, Thomas J., University of Utah, USA.
- organization
- publishing date
- 2025-11-07
- type
- Thesis
- publication status
- published
- subject
- keywords
- wildfire, evacuation, traffic dynamics, driving behaviour, smoke, visibility, virtual reality, traffic simulation, wildland-urban interface
- in
- Report / Department of Fire Safety Engineering, Lund Institute of Technology, Lund University
- issue
- 1076
- pages
- 167 pages
- publisher
- Division of Fire Safety Engineering, Lund University
- defense location
- Lecture Hall V:A, building V, Klas Anshelms väg 14, Faculty of Engineering LTH, Lund University, Lund. The dissertation will be live streamed, but part of the premises is to be excluded from the live stream. Zoom: https://lu-se.zoom.us/s/69763852633
- defense date
- 2025-12-04 09:00:00
- ISSN
- 1402-3504
- ISBN
- 978-91-8104-699-1
- 978-91-8104-700-4
- language
- English
- LU publication?
- yes
- id
- 49e777d7-f159-45b6-9a3c-ada8efad7fb4
- date added to LUP
- 2025-11-07 12:01:40
- date last changed
- 2025-11-11 08:42:45
@phdthesis{49e777d7-f159-45b6-9a3c-ada8efad7fb4,
abstract = {{Introduction: Wildfires pose escalating risks to communities at the wildland--urban interface, often necessitating extensive evacuations. This thesis explores how driving behaviour during wildfire evacuations differs from routine conditions and how these differences can be modelled to improve traffic simulations, thereby supporting more informed planning and response decisions by authorities.<br/><br/>Objectives: Three main research objectives are addressed: (1) characterising how macroscopic traffic dynamics differ between wildfire evacuations and routine traffic conditions, (2) assessing the impact of wildfire smoke on car-following behaviour, and (3) developing a framework for traffic simulations that can support emergency planning for and response during wildfire evacuations.<br/><br/>Methods and outcomes: (1) A dedicated data analysis method was developed to compare traffic dynamics during routine and evacuation scenarios, using traffic detector data from recent wildfire events in California. The analysis revealed that drivers move more slowly and leave larger gaps between vehicles during evacuations. If unaccounted for in simulations, these behavioural shifts can lead to overly optimistic evacuation time estimates.<br/><br/>(2) A custom driving simulator and virtual reality environment were designed to assess how reduced visibility from wildfire smoke affects driver behaviour. Results indicate that, under reduced visibility, drivers reduce their speed when travelling alone, but do not consistently adjust their following distance in congested traffic. These findings inform a visibility-sensitive car-following model.<br/><br/>(3) Finally, a simulation framework was proposed, combining data from real wildfire events, virtual reality experiments and evacuation drills. This framework supports the evaluation of evacuation strategies and planning decisions under varying conditions. A case study applied to a community of more than a thousand households in Colorado demonstrated the framework’s utility in assessing traffic management interventions.<br/><br/>Conclusion: By capturing evacuation-specific driving behaviours and their impact on traffic, this thesis provides practical approaches to enhance the realism of evacuation models, which can, in turn, support more reliable planning and safer wildfire evacuations.}},
author = {{Rohaert, Arthur}},
isbn = {{978-91-8104-699-1}},
issn = {{1402-3504}},
keywords = {{wildfire; evacuation; traffic dynamics; driving behaviour; smoke; visibility; virtual reality; traffic simulation; wildland-urban interface}},
language = {{eng}},
month = {{11}},
number = {{1076}},
publisher = {{Division of Fire Safety Engineering, Lund University}},
school = {{Lund University}},
series = {{Report / Department of Fire Safety Engineering, Lund Institute of Technology, Lund University}},
title = {{Driving behaviour during wildfire evacuation}},
url = {{https://lup.lub.lu.se/search/files/232404264/Driving_behaviour_during_wildfire_evacuation_-_Arthur_Rohaert_-_Doctoral_Dissertation_-_Without_Papers.pdf}},
year = {{2025}},
}