Amazon forest dynamics from multi-temporal airborne lidar data

Douglas Morton, NASA GSFC, douglas.morton@nasa.gov (Presenter)
Veronika Leitold, NASA GSFC, vleitold@gmail.com
Bruce Cook, NASA GSFC, bruce.cook@nasa.gov
Maiza Nara dos Santos, Embrapa, maizanara@gmail.com
Maria Hunter, UNH, mhunter@kaos.sr.unh.edu
Hyeungu Choi, NASA GSFC, hyeungu.choi1@nasa.gov
Marcos Longo, Embrapa, mdplongo@gmail.com
Michael Keller, USDA Forest Service, mkeller.co2@gmail.com

Small footprint airborne lidar data provide unprecedented detail regarding the 3D structure of tropical forests. Canopy characteristics can be resolved at the individual tree scale for quantitative analyses of canopy trees and gaps. Here, we used multi-temporal airborne lidar data for eight Amazon forest sites, each approximately 1,000 ha, to estimate key characteristics of forest disturbance and recovery. Initial lidar collections indicated large differences in forest structure across study sites, including variability in mean canopy height and vertical differentiation of canopy trees. At the 1000 ha site scale, multi-temporal lidar data indicated that canopy structure was stable over short time scales (1 month to 4 years) for intact Amazon forests, with nearly identical canopy height distributions between lidar collections. Compensating processes of canopy turnover, gap formation, and height growth across each site therefore maintained initial differences in canopy structure across study sites in the southern Brazilian Amazon. At the tree scale, we estimated rates of canopy turnover based on evidence for losses in canopy height (≥5 m) between lidar collections. Annualized estimates of canopy turnover ranged between 1-3% for intact forest areas. Canopy losses were further divided into tree fall events (whole tree) and branch fall events (partial crown losses), and gap-forming and non gap-forming events. Overall, canopy losses exceeded new gap formation at all sites, indicating that non gap-forming disturbances constitute an important fraction of all canopy turnover events. Quantitative information from multi-temporal airborne lidar data offers key insights on the regional variability in Amazon forest dynamics. Lidar-based estimates of Amazon forest disturbance and recovery processes over large areas (1,000 ha) complement data from forest inventory plots (1 ha) for improving ecosystem models.

Presentation Type:  Poster

Session:  General Contributions   (Tue 4:35 PM)

Associated Project(s): 

• Morton, Doug: Dynamics of Amazon Forest Disturbance and Recovery from Multi-temporal Airborne LiDAR ...details

Poster Location ID: 180