Repository logo

Browsing by Author "Ney, Jacob, author"

Now showing 1 - 1 of 1
  • Thumbnail Image
    ItemOpen Access
    Comparing crown fire predictions in ponderosa pine stands among four fire behavior models
    (Colorado State University. Libraries, 2024) Ney, Jacob, author; Hoffman, Chad, advisor; Linn, Rodman, committee member; Fischer, Emily, committee member
    Fire and land managers commonly use fire behavior modeling systems to support their planning and decision-making process. Fire modeling systems have been increasingly used across the western United States to plan fuel treatments that reduce hazard fuels, especially as a drier climate has resulted in more frequent high severity wildfire. Given differences in model types, approaches, assumptions, and sensitivity to various input parameters, modeling systems can produce different predictions and lead to different management decisions. Variability arising from model selection results in increased uncertainty within the decision-making framework. Multi-model comparisons help identify areas of model agreement and disagreement, reduce uncertainty associated with management decisions, and identify directions for future experimentation. Here, I compare predictions of fire type and crown fire rate of spread (ROS) among four modeling systems that represent a range of model types and complexities—Wildland-urban interface Fire Dynamics Simulator (WFDS), QUIC-Fire, a Rothermel-based modeling framework, and Crown Fire Initiation and Spread (CFIS). Comparisons (n = 297) were made based on a range of forest structure and environmental conditions representative of treated and untreated ponderosa pine forest stands in the southern Rocky Mountains. All four models predicted crown fire occurrence for 71% of simulations in total. WFDS, QUIC-Fire, and CFIS agreed on fire type more than 65% of the time. Rothermel predicted crown fire for 41% of simulations with ROS predictions 45% lower than the other models. Models tended to agree on crown fire occurrence in scenarios with a low canopy base height and greater surface and canopy fuel loading, indicating lower uncertainty in predicted fire behavior among models when fuel hazard is greatest. Differences among model predictions were more evident in scenarios with greater canopy base heights, moderate surface and canopy fuel levels, and at lower windspeeds. These results suggest that uncertainty introduced by model selection is likely greatest for designing and evaluation of fuel treatments, and that further research on fire behavior in treated forests stands is needed.