OOS 9-6 - Establishing the cross-compatibility of aerial and terrestrial LiDAR systems for quantifying forest structure

Tuesday, August 13, 2019: 3:20 PM
M107, Kentucky International Convention Center
Franklin Wagner1, Elizabeth A. LaRue1, Jeffrey Atkins2, Brady Hardiman1 and Songlin Fei1, (1)Forestry and Natural Resources, Purdue University, West Lafayette, IN, (2)Department of Biology, Virginia Commonwealth University, Richmond, VA
Background/Question/Methods

Forest canopy structure (CS) strongly influences a wide variety of ecosystem functions, including productivity, precipitation and gas exchange, nutrient cycling, and habitat provisioning. However, accurately measuring the 3-D structural complexity of canopies remains a challenge and a topic of interest in forest ecology. LiDAR is a remote sensing technology with demonstrated potential to measure 3-D CS. Generally, the higher measurement density of terrestrial LiDAR systems enables much finer resolution of CS than aerial systems, but these lack the spatial extent of aerial systems. Establishing the cross-compatibility of the two could provide a means of extrapolating the detail achievable with terrestrial systems to the extent attainable with aerial systems. Characterizing canopy structure in detail at larger spatial scales would facilitate incorporation with other remote sensing data and climatic models, and could provide important information to support forest management decisions. Our objective was to investigate the relationships between suites of CS metrics derived terrestrially and aerially at seven National Ecological Observatory Network sites. This was done by evaluating the correlation strength between the various CS metrics.

Results/Conclusions

Relationships between the LiDAR derived metrics varied by category (height, leaf area, cover, and heterogeneity). Maximum canopy height and mean outer canopy height showed correlations of r = 0.94 and r = 0.84, respectively, suggesting that these are measured similarly between both systems. Mean vegetation area index showed a correlation of r = 0.88, providing evidence that this metric is measured similarly. This is encouraging as vegetation area is a critical component of ecosystem functioning. Canopy gap and cover fraction showed values of r = 0.81 and r = 0.72, suggesting that the two LiDAR systems are effective at evaluating canopy gapiness—an important measure related to canopy light use efficiency. Metrics related to canopy heterogeneity showed a range of r = 0.00-0.75, indicating that evaluating the 3-D structural complexity of canopies via LiDAR may still be difficult. Overall, these results suggest the feasibility of extrapolating the detail of terrestrial LiDAR to the scale of aerial LiDAR for most measures. However, evaluating the structural complexity of canopies at large extents may still be a challenge that depends on a variety of factors such as forest type, LiDAR systems used, and resulting point densities.