Characterizing fuelbed structure, depth, and mass in a grassland using terrestrial laser scanning

Fuelbed depth and loading are fundamental parameters for predicting fire behavior and effects. This thesis considers the application of Terrestrial Laser Scanning (TLS) for characterizing the structure, measuring depth and mass in a grassland fuelbed. It examines whether TLS can be used to quantify the vertical structure of a bunchgrass community in terms of bunch and seed head, and to describe spatial variability in height and fuel mass at fine grain (0.25m2). In the experiment, fuels are mechanically manipulated to enhance variability. Height and mass are modified independently at three levels using a randomized design. Results show that vertical strata (bunches and seed head) are measurable within the TLS height profile and changes in height due to treatments can be detected. There are statistically significant but small absolute differences (± 1-6cm) between TLS-derived and field-measured heights with no systematic bias observed. TLS-derived height measurements have a higher precision than field-measured heights, but the accuracy of measurements is uncertain given ambiguity in field measurements. In the untreated grassland, fuel mass is associated with TLS-derived bunch height and standard deviation of height; a linear model using these metrics explains ~30 percent of the variability in biomass. In the treated grassland, a linear model using median height and standard deviation of height accounts for ~40 percent of the variability in biomass. In the biomass-only treatment, bunch height and canopy cover are the best combination of explanatory variables for biomass accounting for ~ 42 percent; median height and standard deviation of height account for 24 percent of biomass variability in the height-only treatment. Collectively, these results suggest that TLS can be used to quantify the spatial variability in heights (and hence, volumes) occupied by fuels in a grassland where biomass is concentrated in distinct strata. However, prediction of biomass leaves room for improvement, with the most substantial gains likely to be made with better cover/density metrics. Even with improved density metrics, improvements are likely to be modest because variability in the bunchgrass system is low and the TLS is not very sensitive to small changes in cover.

Name thesis:degree_name

Master of Science (MS)

Grantor dc:publisher

University of Montana

Year

2013

Author dc:creator

• Umphries, Tara Amber

• fuels
• grassland
• Lidar

Repository record dc:identifier

https://scholarworks.umt.edu/etd/188

OAI identifier oai:identifier

oai:scholarworks.umt.edu:etd-1207