Year of Award

2013

Document Type

Thesis

Degree Type

Master of Science (MS)

Degree Name

Forestry

Department or School/College

College of Forestry and Conservation

Committee Chair

Carl Seielstad

Commitee Members

David Affleck, Ragan Callaway

Keywords

3D, conifer, crown, laser, LiDAR, profile, Ripley's K, TLS, Weibull

Publisher

University of Montana

Abstract

Emerging interests in wildland fire behavior and risk, bioenergy utilization, carbon sequestration, and wildlife conservation increasingly rely on accurate assessments of the amount and location of biomass within the dominant plants on the landscape, often at finer scales than traditional methods have provided. At the tree scale, current studies often distribute biomass uniformly through simple volumes (e.g., cones and cylinders). However, biomass is heterogeneous at a variety of scales from needle clusters to groups of trees. This thesis presents techniques for using terrestrial laser scanning data to define crown profiles and describe within-crown heterogeneity in Pseudotusga menziesii, Pinus ponderosa, and Abies lasiocarpa of the Interior Northwest. Crown profiles were modeled using parametric curves applied to crown-length normalized laser point clouds, dimensioned by height above ground and distance from bole-centroids. A crown-base metric was derived from the laser data and compared to conventional field measurements. For all species, a modified Weibull curve fit crown points with significantly smaller error than a beta curve, cone, or cylinder; crown profile Weibull curves were species-specific and not interchangeable without producing signifcantly greater error. Within-crown patterning was described using a 3-D form of the Ripley’s K function. Ripley’s K analysis detected maximum clustering occuring at scales of 1.25 – 2.50 percent of crown length (e.g., 25-50 cm radius clusters in a 20 meter crown). P. ponderosa demonstrated clustering over the largest range of scales and to the greatest degree, while A. lasiocarpa exhibited clustering over the smallest range of scales. The scale of clustering did not change when points roughly corresponding to branchwood were excluded from the analysis. This study provides groundwork for predicting the spatial distribution of biomass with tree crowns. Limitations of the work include uncertainty regarding the impacts of occlusion of inner crowns and the relationships between laser points and foliage-branch elements, and the lack of spatial explicitness inherent to Ripley’s K. Future work should examine these issues with an eye toward refinement of predictive models linking traditional biomass allometry with spatial arrangement of canopy material.

Share

COinS
 

© Copyright 2013 Jena Ferrarese