Final report: Applying multispectral 3D aerial imagery in the Los Amigos Reserve in eastern Peru
2006
Report
20
Unpublished report of Winrock International (WI) for the Gordon and Betty Moore Foundation
Turners Falls, USA
Los Amigos;remote sensing;intermediate spatial scales;carbon;biomass;vegetation;threats;impacts of economic activities
Digital aerial imagery offers breakthrough potential to improve landscape scale conservation science and threat detection for long-term protection of conservation reserves. This project has resulted in proof of concept for the application and use of digital aerial imagery to support conservation and conservation science in the Los Amigos Conservation Concession (LACC), Madre de Dios, Peru and the Osa Peninsula, Costa Rica. A major advance the project has made is in the ability to move from small fixed plot analysis to the landscape scale. The two forest carbon inventories (LACC and the Osa Peninsula) conducted through this project demonstrate the value of integrating large sample areas (in the form of aerial transects) with traditional forest inventory ground plots. The LACC inventory notes that the reserve contains over 79 million tons of CO2 in the above-ground biomass – roughly equivalent to the projected 2006 diesel and jet fuel emissions from the State of California. The process of doing this inventory helped refine automated methods of biomass estimation, which makes it possible to classify thousands of hectares per person-day, compared to tens of hectares using manual methods. This provides an opportunity to understand both natural and human-induced change on larger areas in a cost-effective way. The modeling of future threats to LACC demonstrated the direction and potential timing of deforestation pressure that could affect the reserve. This analysis drew upon existing research and provided additional ideas on how and where to mitigate these threats well in advance of the deforestation front. Multi-temporal aerial images can then track the disappearance of or changes in individual trees and clearly demonstrate a cost-effective means of monitoring illegal logging within the reserve and begin to measure time-dependent natural phenomena such as the natural mortality (Appendix 3). Improved efficiencies gained in this process have reduced data collection and processing costs from US$1.15 per ha in 2003 to approximately US$0.12 per hectare at a project scale of approximately 300,000 ha or more. Applications of the data have made it possible to classify and estimate forest values such as biomass and forest carbon in areas of the reserve that have not been visited by LACC staff or researchers. Scaling-up the lessons learned from activities in this project would enhance long-term conservation potential and expanded conservation science across the Amazon Basin by working together with a network of organizations in selectively targeting these technologies and methods to areas of greatest need.