Graduate Thesis Or Dissertation
 

Modeling the Influence of Wildfire Management Networks in Social-Ecological Systems

Public Deposited

Downloadable Content

Download PDF
https://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/mk61rr43p

Descriptions

Attribute NameValues
Creator
Abstract
  • Wildfire management networks allow for collaborative intervention in mitigating the ever-present risk of major wildfire events. This thesis consists of two manuscripts that explore the topology and characteristics of wildfire management networks and their ability to influence wildfire severity in social-ecological systems. The first manuscript focuses on the characteristics of wildfire management networks and their ability to transmit signals of risk. The second manuscript explores the wildfire management network’s influence on the landscape and the trade-off between investing in treatments on the landscape vs. investing in the network’s ability to guide treatments. The first manuscript combines analytical techniques from fields like graph theory, sociology, and machine learning to identify critical components in a successful, influential wildfire management network. A social network influence propagation model was used to transmit signals of risk from the landscape through the management network to actors on the landscape. The model was first run on a test set of simulated random networks before being run on a real-world network. Multiple machine learning algorithms were used to identify which network metrics, network motifs, and topological characteristics were most important for a network to generate a high target response variable, network transmission efficiency (NTE). Our results suggest that we can predict a network will have low NTE if the network has a low edge density between the landscape and the network as well as within the network itself. Our simulations predict a network will have high NTE if it has a high network edge density, many nodes, and many connections to landscape actors. Common network characteristics that high NTE networks share include high betweenness centrality, high connectivity to the landscape signal and landscape actors, and a high average degree centrality. These factors were also found to exist in a successful real-world wildfire management network with the connectivity to the landscape actors being the most important variable in raising NTE. The results suggest that there are optimal wildfire management networks configurations that maximize their influence on the landscape The second manuscript explores the impact the wildfire management network has in a coupled social-ecological system. Envision in an agent-based landscape futures simulation model used to explore the effects of various management scenarios over time. The decision profiles of landscape actors in Envision were modified so that the influence of the wildfire management network shaped the landscape actor’s decision alongside the original variables that governed their actions. Twenty-two, 40-year wildfire scenarios were run in Envision to test how sensitive the landscape actors are the wildfire management network’s response variable in the first manuscript, NTE, and the landscape actor’s sensitivity to different levels of investment in the landscape vs. investment in the network. We found that the most important variables from the first paper reduced the NTE significantly in these runs. However, the number of houses burned on the landscape could not be predicted nor shared a relationship with the NTE or cost of the network. The amount of land treated on the ground shared no relationship with a reduction in the number of houses burned, which suggests wildfires did not happen in areas of our simulation that the wildfire management networks could assist in treating. These model outcomes suggest that future work needs to explore how to the tradeoffs between the costs of collaboration versus fuels reduction and how to model that relationship.
  • Keywords: Network Science, Environmental Modeling, Social Influence Networks, Social Networks
License
Resource Type
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Academic Affiliation
Rights Statement
Publisher
Peer Reviewed
Language
Embargo reason
  • Pending Publication
Embargo date range
  • 2023-07-29 to 2024-02-29

Relationships

Parents:

This work has no parents.

In Collection:

Items

Thumbnail Title Date Uploaded Visibility Actions
file details AvilaColton2024.pdf 2023-08-07 Public Download