Abstract

Context: The Great Basin in the western US features arid to semi-arid climates and sagebrush steppe ecosystems. Increased fire frequency can negatively impact these ecosystems by promoting invasive species. Agent-based modeling is a valuable tool to study fire dynamics in sagebrush steppe ecosystems, offering insights in management strategies and ecosystem resilience. Objectives: We developed SageFire, an agent-based model, to simulate heat dynamics influenced by shrub cover during fires in the Great Basin. Our objectives were to accurately classify ground cover types, incorporate classified imagery into a simulation model, and accurately visualize fire intensity based on the spatial arrangement of the three ground cover types. Methods: SageFire is based on the Onaqui Flat SageSTEP site in western Utah, which represents a low-elevation Wyoming big sagebrush ecosystem. SageFire was built in NetLogo and simulates a fire event by considering ground cover types (determined using aerial imagery that was classified into ground cover types using eCognition software) and proximity to shrub centers. Results: The imagery classification for high, medium, and low shrub scenarios had kappa values of 0.78, 0.86, and 0.77, respectively. The SageFire model successfully incorporates classified imagery and demonstrates the impact of vegetation arrangement on fire intensity. Conclusion: SageFire presents a valuable tool for land managers in the Great Basin, offering insights into the relationships between vegetation distribution and fire intensity in Wyoming big sagebrush landscapes. The model's customizability facilitates location-specific decisions, enabling land managers to evaluate potential fire intensity levels and subsequent perennial bunchgrass mortality, which is instrumental in prioritizing preemptive restoration treatments.

Degree

College and Department

Life Sciences; Plant and Wildlife Sciences

Rights

https://lib.byu.edu/about/copyright/