Light detection and ranging (LiDAR) : what we can and cannot see in the forest for the trees

Graduation date: 2011Recently concerns over anthropogenic carbon pollution have received increased\ud global attention and research in forest biomass and carbon sequestration has gained\ud momentum. Light Detection and Ranging (LiDAR) remote sensing has in the last\ud decade demonstrated forest measurement and biomass estimation potential. The\ud project objective was to compare LiDAR forest biomass estimates to traditional field\ud biomass estimates in a conifer predominant forest located in the Pacific Northwest\ud region of the United States.\ud Chapter 2 of this dissertation investigated mapping-grade GPS accuracy in\ud determining tree locations. Results indicated that post processing of coded\ud pseudorange satellite signals is the most accurate of those we tested for GPS surveying\ud under a conifer dominant forest canopy. Chapter 3 compared LiDAR, total station, and\ud GPS receiver discrete point elevations and DEMs across a range of forest settings.\ud Average total station plot elevation differences ranged from -0.06 m (SD 0.40) to -0.59 m (SD\ud 0.23) indicating that LiDAR elevations are higher than actual elevations. Average plot GPS\ud elevation differences ranged from 0.24 (SD 1.55) to 2.82 m (SD 4.58), and from 0.27 (SD\ud 2.33) to 2.69 m (SD 5.06) for LiDAR DEMs.\ud Chapter 4 assessed LiDAR’s ability to measure three-dimensional forest structure\ud and estimate biomass using single stem (trees and shrubs) remote sensing. The LiDAR\ud data tree extraction computer software programs FUSION, TreeVaW, and watershed\ud segmentation were compared. LiDAR spatial accuracy assessment resulted in overall\ud average error and standard deviation (SD) for FUSION, TreeVaW, and watershed\ud segmentation of 2.05 m (SD 1.67 m), 2.19 m (SD 1.83 m), and 2.31 m (SD 1.94 m)\ud respectively. Overall average LiDAR tree height error and standard deviations (SD)\ud respectively for FUSION, TreeVaW and watershed segmentation were -0.09 m (SD\ud 2.43 m), 0.28 m (SD 1.86 m), and 0.22 m (2.45 m) in even-age, uneven-age, and old\ud growth plots combined; and for one clearcut plot 0.56 m (SD 1.07 m), 0.28 m (SD\ud 1.69 m), and 1.17 m (SD 0.68 m), respectively. Biomass comparisons included feature\ud totals per plot, mean biomass per feature by plot, and total biomass by plot for each\ud extraction method. Overall LiDAR biomass estimations resulted in FUSION and\ud TreeVaW underestimating by 25 and 31% respectively, and watershed segmentation\ud overestimating by approximately 10%. LiDAR biomass underestimation occurred in\ud 66% and overestimation occurred in 34% of the plot comparisons