OOS 68-6
Biomass monitoring in the western Amazon: Using unmanned aircraft to provide near real-time updates of biomass estimates in a conservation hotspot

Thursday, August 13, 2015: 9:50 AM
341, Baltimore Convention Center
Maxwell C. Messinger, Biology, Wake Forest University, Winston-Salem, NC
Miles R. Silman, Biology, Wake Forest University, Winston-Salem, NC
Gregory P. Asner, Department of Global Ecology, Carnegie Institution for Science, Stanford, CA
Marcus W. Wright, Center for Energy, Environment, and Sustainability, Wake Forest University, Winston-Salem, NC
William Nicholson, Biology, Wake Forest University, Winston-Salem, NC
Background/Question/Methods

Small-unmanned aerial systems (SUAS) can provide new ways to observe the environment and replace expensive or labor-intensive inventory methods. Forest carbon is a key uncertainty in the global carbon cycle and also important for carbon conservation schemes. Using SUAS, we developed methods for frequent, low-cost forest carbon monitoring to reduce this uncertainty. State-of-the-art techniques in forest carbon monitoring typically involve the use of costly manned aircraft and sensors to estimate aboveground biomass (AGB). While these techniques produce accurate estimates of AGB, regularly updating estimates is cost-prohibitive and, as a result, larger-scale satellite-derived estimates of deforestation and reforestation are used to make these updates. Here we report results from a series of missions to estimate forest structure and biomass in the Andes and Amazon of Perú. We sought to: (1) determine the feasibility of operating fixed-wing SUAS unsupported in remote areas for extended periods; and (2) identify a robust method for frequent and accurate AGB estimation. Using SUAS we collected imagery of tropical rainforest in the western Amazon basin on scales ranging from single ha to tens of km2and at pixel resolutions ranging from 5-20cm.

Results/Conclusions

Using structure-from-motion (SFM) techniques we created 3-dimensional models of forest canopy. Combining SUAS data with existing active remote sensing data, we estimated forest AGB using two methods. First, we used a previously established digital elevation model (DEM) and the current SFM data to derive tree and canopy heights, and then used wood density data and allometric equations to estimate AGB. Second, we derived a biomass estimate using a novel methodology based on large canopy trees. Together, the estimates provide a significant improvement in AGB estimation through time as compared to simple estimation of deforestation and reforestation. We also discuss challenges associated with SUAS operation in remote areas and methods derived to overcome them. Cheap, easy to use, and with the ability to collect high-quality spatial data, small-unmanned aircraft provide researchers and land managers with a new ability to detect and monitor changes to AGB in near real-time. We discuss implications and opportunities in the study, management, and protection of forest carbon reserves.