Monitoring Forest Carbon Dynamics Using Echidna@ Ground-based LIDAR

The high geometric accuracy of Lidar data provides the opportunity to fully quantify forest structure and forest carbon dynamics associated with deforestation and natural disturbance. The main objective of this research is to measure forest structure, biomass and carbon dynamics in selected New England forest conifer stands over a 2-year time period using a ground-based, near-infrared (1064 nm), full-waveform lidar, the Echidna® Validation Instrument (EVI).

Forest canopy structure was measured using the EVI at two sites in New England: the Harvard Forest Hemlock site (Massachusetts) and the Howland Forest Shelterwood site (Maine). To assess change during the two-year period, structure and biomass parameters, including mean tree diameter at breast height, stem count density, basal area, and above-ground biomass, were retrieved with the lidar in 1-ha plots at each site in 2007 and again in 2009. In prior work, we validated lidar retrievals from these and similar stands using conventional measurement techniques, obtaining R2 values of 0.84, 0.97, 0.96 and 0.98, respectively, for the parameters listed above. The Harvard Forest Hemlock site, a hemlock plantation stand, suffered significant damage to woody biomass and loss of leaf area during an ice storm in December 2008 that broke branches, felled crowns, and downed trees throughout the forest. Site-level EVI-retrieved leaf area index (LAI) was 0.57 m2/m2 lower and retrieved above-ground biomass was 12 Mg/ha smaller in the Hemlock site from 2007 to 2009. The Howland Forest Shelterwood site, a natural stand of mixed conifer forest, was thinned in spots and patches around 1998 to remove about one-third of its biomass, releasing many trees from light competition. Here, the values of biomass and leaf area increased as the stand grew without disturbance from 2007 to 2009. Site-level EVI-retrieved leaf area index (LAI) was 0.13 m2/m2 higher and retrieved above-ground biomass was 13 Mg/ha larger in the Shelterwood site from 2007 to 2009. A canopy height model (CHM), built from EVI point cloud data, provided detailed tree height information at each site, and a comparison of CHMs showed detailed change in forest height, including the identification of missing crowns. The spatial distribution of canopy height change matched the field observations. At the Hemlock site, the apparent decrease of canopy height showed the location of trees and crowns felled by the weight of the ice. At the Shelterwood site, the apparent increase of canopy height showed the locations of fast growing trees.

The fine spatial resolution of the EVI data allows biomass and biomass change estimates to be derived at hectare or sub-hectare levels. By quantifying the uncertainty of EVI-retrived biomass changes, this approach can be used to validate products of airborne and spaceborne remote sensors focused on biomass dynamics, such as MODIS and ICESat-2.