Evaluating Tree Structural Allometric Variability with High Resolution LiDAR Data

Related Research Areas
Carbon Cycle & Ecosystems
Project Description
Forest ecosystems remain one of the largest terrestrial carbon stocks on Earth, but the both the stock and flux of this critical carbon pool remain unknown. Developing a baseline map of global carbon stocks in forests has been a research initiative of great interest in the remote sensing community. Remote sensing-based aboveground biomass (AGBM) maps are developed by fitting RS data to field-based estimates of AGBM. Unfortunately, considerable error remains in the field estimates due the application of allometric equations relating individual tree structure to individual biomass. These allometric equations have been developed from relatively small samples of tree populations, and therefore may not be applicable across varying environments or life histories. This research aims to determine ranges of tree allometric variability by exploring how individual tree height, crown radius and crown area vary across many ecosystems in the US, Canada, and the tropics. High resolution LiDAR datasets are run through an efficient 3D crown delineation algorithm. The algorithm has been developed and is run through the NASA Advanced Supercomputing facility to produce estimates of individual tree location, height, and crown dimensions across entire landscapes. Allometries between tree heights and crown dimensions will be mapped in comparison to forest age since disturbance (from landsat) and environmental attributes (climate, topography) to determine the ranges within which it is appropriate to apply a single allometric equation for a given tree species. This research should inform both the forestry and remote sensing communities as to where field-based maps may have high errors and more biomass-based allometries are required.
Project Administrator(s):
Laura Duncanson


Laura Duncanson