Effects of stand attributes in evaluating even-aged loblolly pine volume with LiDAR

The purpose of this project is to determine the practicality of using LiDAR technology in the field as a primary tool for forest inventory. Specifically, this project uses variables generated from Quality Level 2 (QL2) LiDAR data obtained from the state of North Carolina Floodplain Mapping Program to investigate the overall goodness of fit of Loblolly pine volume based on ground measurements. Seven stands in the Piedmont and coastal regions in North Carolina were used in this study. Volume was calculated using pre-harvest cruise data and was regressed against the LiDAR generated predictors of height and canopy cover. The analysis was conducted on both plot level and at a mid-sized management level, of theoretical harvest units (THUs). These LiDAR derived variables were only moderately successful at estimating loblolly pine volume at the plot level (R2 =0.45) and at the THU level (R2 = 0.37). When ground calculated height was regressed against LiDAR estimated height, no bias was detected indicating that while model fit was modest, the overall approach was correct.