Unmanned aerial systems for forest structure mapping: assessments of area-based and individual tree monitoring
Date
2020
Authors
Creasy, Matthew, author
Tinkham, Wade T., advisor
Vogeler, Jody, committee member
Hoffman, Chad, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Characterization of forest structure is important for management-related decision making, especially in the wake of disturbance. Increasingly, observations of forest structure are needed at both finer resolution and across greater extents in order to support managers in meeting spatially explicit objectives. Current methods of acquiring forest measurements can be limited by a combination of time, expense, and either extent or temporal resolution. Drone or UAS-based photogrammetry provides an airborne method of forest structure data acquisition at a significantly lower cost and time commitment when compared to existing methods of such as airborne laser scanning (LiDAR). A growing body of literature confirms UAS-based photogrammetry models can be as detailed as conventional LiDAR models. However, there exists a knowledge gap in best practice for data acquisition parameters and assessment of accurate characterization within forest photogrammetry. The following two chapters utilize large stem mapped sites to fill that knowledge gap by 1) systematically testing the effects of UAS flight speed and altitude on plot-based aboveground biomass modeling through photogrammetry and 2) evaluating several algorithms for detecting individual tree locations and characterizing crown areas. Results show a strong positive relationship between flight altitude and aboveground biomass modeling, with all UAS flights evaluated above 80 m altitude, providing better results (2-24% more variance explained) than contemporary LiDAR modeling strategies. Additionally, results demonstrate that the probability of detecting individual trees decays moving from the dominant overstory to suppressed trees, corresponding to >97% at the top of the canopy and decreasing to 67% for trees in the understory. Our results indicate the potential for UAS photogrammetry to produce highly detailed maps of forest biomass, as well as capture variation of forest structure through the detection of trees and tree groups. Such high-resolution data has the potential to become a much-needed tool for monitoring forest structures to inform spatially explicit management objectives. Additionally, these studies reinforced how UAS photogrammetry can provide low-cost repeat monitoring of forest conditions.
Description
Rights Access
Subject
drone
photogrammetry
UAS
individual tree detection
aboveground biomass
structure from motion