UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Stand structure classification, succession, and mapping using LiDAR Moss, Ian


In this dissertation, a consistent, reasonably precise, verifiable system of stand structure classification was developed and demonstrated. The goal was to provide a foundation for better communication amongst forest management professionals. A novel distance metric and classification algorithm were introduced. The distance metric was based on similarity in reversed cumulative stems and basal area per ha by diameter (DBH; 1.3 m above ground). This distance metric: (1) uses commonly available information; (2) avoids the separation of data into arbitrary DBH classes; and (3) represents a broad range of simple to complex stand structures. Using 421 plots established across a range of Interior Douglas-fir (Pseudotsuga menziesii var. glauca (Beissn.) Franco) and lodgepole pine (Pinus contorta var. latifolia (Engelm.) Critchfield) stands in the Cariboo region of British Columbia, Canada, a 17-class system of classification was constructed. Whole stand statistics, cumulative distributions, and stand structure/distribution indices were used to evaluate the results. The classes were reasonably precise, with meaningful partitions separating single layered versus complex stands. The utility of the classification system was investigated for diagnosing potential patterns of succession. Over 100 simulated stand structure progressions were simulated using plot data input into an individual-tree growth model. Similar progressions in stand structure classes were assigned common pathways. Four general patterns of succession were observed: (1) a high density single layered pathway; (2) a moderate density single layered pathway; (3) a moderate density complex pathway; and (4) a moderate density, mixed complex-single layered pathway. Lastly, the feasibility of using aerial Light Detection and Ranging (LiDAR) for stand structure classification in forest inventory was assessed. LiDAR was reasonably effective in distinguishing structural classes on the basis of cumulative distributions in basal area or gross volume with respect to DBH, but it was less successful when the distributions in numbers of stems per ha were included. Further study using additional LiDAR metrics beyond those used in this study are needed to improve the use of LiDAR for stand structure classification. This stand structure classification system has potential for a wide variety of forest management applications, including improvement of linkages between strategic and tactical planning and implementation.

Item Media

Item Citations and Data


Attribution 3.0 Unported