Using LiDAR Data to Estimate Effective Leaf Area Index, Determine Biometrics and Visualize Canopy Structure in a Central Oregon Forest with Complex Terrain

Leaf Area Index (LAI), the total one-sided area of leaf tissue per unit ground surface area, is an important parameter in many ecological models. LAI is important for determining interception loss, and can be used potentially as a surrogate for other ecosystem parameters when studying ecosystem processes and services. Estimation of LAI at the watershed scale is difficult since traditional, direct destructive methods are cumbersome and possible only on small spatial scales. Furthermore, estimation of LAI in steep terrain has proven challenging for indirect methods using tools that utilize lasers to estimate light penetration through canopies. In this study, digital hemispherical photographs were used to ground-truth a Light Detecting and Ranging (LiDAR) method of estimating effective LAI at both the plot and watershed scales using canopy volume from LiDAR point cloud data. Effective LAI differs from true LAI in that it includes non-leaf material, such as branches, in the calculation. The LiDAR model seems to underestimate effective LAI when compared to ground based methods (R2= 0.3346, p<.001 for="" of="" the="" vegetation="" plots="" in="" watershed="" h.j.="" andrews="" experimental="" forest.="">

LiDAR data were also used to calculate biometrics (height, crown diameter, and stem location) of individual trees and to visualize forest structure. When compared to vegetation surveys completed for all permanent vegetation plots, 82% of live trees were identified using LiDAR data. The results of this work can be used for modeling throughfall, canopy storage and interception loss for the watershed, either scaling from branch, to plot, to watershed leaf area or using allometric equations with the identified individual trees. The visualizations presented could assist researchers by allowing them to see gaps in the canopy and assess variability on a subplot scale. Future research includes assessing what factors affect the accuracy of tree identification and how software programs can be improved for more accurate tree identification and LAI estimation in complex terrain.