Estimating Basal Area and Stem Volume for Individual Trees from Lidar Data

This study proposes a new metric called canopy geometric volume G, which is derived from small-footprint lidar data, for estimating individual-tree basal area and stem volume. Based on the plant allometry relationship, we found that basal area B is exponentially related to G (B 1G, where 1 is a constant) and stem volume V is proportional to G (V 2G, where 2 is a constant). The models based on these relationships were compared with a number of models based on tree height and/or crown diameter. The models were tested over individual trees in a deciduous oak woodland in California in the case that individual tree crowns are either correctly or incorrectly segmented. When trees are incorrectly segmented, the theoretical model B 1G has the best performance (adjusted R2, 0.78) and the model V 2G has the second to the best performance ( 0.78). When trees are correctly segmented, the theoretical models are among the top three models for estimating basal area ( 0.77) and stem volume ( 0.79). Overall, these theoretical models are the best when considering a number of factors such as the performance, the model parsimony, and the sensitivity to errors in tree crown segmentation. Further research is needed to test these models over sites with multiple species. Introduction Accurate forest structural information is crucial to a number of applications including forest management (Maltamo et al., 2004a), fire behavior analysis (Riaño et al., 2004), and global warming and carbon management (Birdsey and Lewis, 2002). Compared to optical remotely sensed imagery, lidar (Light Detection and Ranging) can directly measure the vertical Ra Ra Ra Ra Estimating Basal Area and Stem Volume for Individual Trees from Lidar Data Qi Chen, Peng Gong, Dennis Baldocchi, and Yong Q. Tian canopy information and is gaining popularity in forest and ecological studies (Lefsky et al., 2002a; Hall et al., 2005). With the high pulse-density of small-footprint lidar data, nowadays it is possible to isolate individual trees and directly extract individual-tree locational and dimensional parameters including treetop locations, tree heights, crown sizes, and even crown boundaries (e.g., Popescu et al., 2003; Hyyppä et al., 2001; Persson et al., 2002). We developed a new method to delineate individual tree crowns in a savanna woodland in California using small-footprint lidar data (Chen et al., 2006). This companion study is to further extract individual-tree structural parameters including basal area and stem volume, which cannot be directly measured by lidar, but can be potentially related to tree dimensional information such as tree height and crown size. This study is part of a large project, in which we parameterize an individual-tree based biogeochemical model called “MAESTRA” (Medlyn, 2003) for spatially-explicit ecological modeling. Extracting individual-tree structural parameters from small-footprint lidar data is useful not only for our detailed ecological modeling but also for the large-scale forest inventory. There are at least three kinds of methods to perform the large-scale forest inventory with small-footprint lidar data: the first is called stand-level regression method. This method is to create regression models between canopy structural parameters and the laser pulses statistics at the stand or plot levels (Means et al., 2000; Holmgren et al., 2003; Næsset et al., 2004; Næsset, 2004; Riaño et al., 2004; Hall et al., 2005). The statistics used in the regression models typically include the height mean, minimum, maximum, variance, coefficient of variance, percentiles, and the percentage of canopy returns. The second method is called individual-tree integration method. If the tree dimensional information such as individual tree height and crown size can be accurately extracted, the individual tree canopy structure parameters can be derived based on allometric equations or regression models (Hyyppä, et al., 2001; Persson et al., 2002; Næsset and Økland, 2002; Riaño et al., 2004; Roberts et al., 2005). Then, the canopy structure information over a large area can be obtained by simply integrating individual tree values. Compared to the stand-level regression method, the ground truth canopy structure measurements for developing the regression models are needed only for a sample of trees instead of many of plots or stands, which can significantly reduce the fieldwork. However, not many studies have used PHOTOGRAMMETRIC ENGINEER ING & REMOTE SENS ING Decembe r 2007 1355 Qi Chen is with the Department of Geography, University of Hawaii at Manoa, Honolulu, HI 96822 ([email protected]). Dennis Baldocchi is with the Department of Environmental Science, Policy, and Management, 137 Mulford Hall, University of California at Berkeley, Berkeley, CA 94720 ([email protected]). Peng Gong is with the Center for the Assessment and Monitoring of Forest and Environmental Resources (CAMFER), 137 Mulford Hall, University of California at Berkeley, Berkeley, CA 94720, and the State Key Lab of Remote Sensing Science (jointly sponsored by Institute of Remote Sensing Applications, Chinese Academy of Sciences, and Beijing Normal University), Postal Box 9718, Beijing, P.R. China 100101. Yong Q. Tian is with the Department of Environmental, Earth and Ocean Sciences, University of MassachusettsBoston, Boston, MA 02125. Photogrammetric Engineering & Remote Sensing Vol. 73, No. 12, December 2007, pp. 1355–1365. 0099-1112/07/7312–1355/$3.00/0 © 2007 American Society for Photogrammetry and Remote Sensing 05-131.qxd 2/11/07 10:34 Page 1355