Lidar-derived volume metrics for aboveground biomass estimation in conifer stands

Tao, S., Li, L., Q. Guo, L. Li, B. Xue, M. Kelly, W. Li, G. Xu, and Y. Su. 2014. Airborne Lidar-derived volume metrics for aboveground biomass estimation: A comparative assessment for conifer stands. Agriculture and Forest Management 198–199: 24–3

Estimating aboveground biomass (AGB) is essential to quantify the carbon balance of terrestrial ecosystems, and becomes increasingly important under changing global climate. Volume metrics of individual trees, for example stem volume, have been proven to be strongly correlated to AGB. In this paper, we compared a range of airborne Lidar-derived volume metrics (i.e. stem volume, crown volume under convex hull, and crown volume under Canopy Height Model (CHM)) to estimate AGB. In addition, we evaluated the effect of horizontal crown overlap (which is often neglected in Lidar literature) on the accuracy of AGB estimation by using a hybrid method that combined marker-controlled watershed segmentation and point cloud segmentation algorithms. Our results show that: 1) when the horizontal crown overlap issue was not addressed, models based on point cloud segmentation outperformed models based on marker-controlled watershed segmentation; models using stem volume estimated AGB more accurately than models using crown volume under convex hull and crown volume under CHM. 2) Once the horizontal crown overlap issue was taken into consideration, the model using crown volume under individual trees in the lidar cloudCHM yielded a more accurate estimation of AGB. Our study provides a comprehensive evaluation of the use of airborne Lidar-derived volume metrics for AGB estimation and could help researchers choose the appropriate airborne Lidar-derived volume metric. Moreover, the results also indicate that horizontal crown overlap should be addressed when the airborne Lidar-derived forest crown volume is used for estimating AGB. Pdf download. Journal link.

Keywords: Lidar; aboveground biomass; crown volume; individual tree segmentation