REDD Fast Logging Assessment & Monitoring Environment (REDD-FLAME): Early Detection of Deforestation using SAR
Pearson, Tim¹; Verhoeven, Rob²; Hoekman, Dirk³; Franke, Jonas⁴; van Valkengoed, Eric⁵; van Roosmalen, Vasco⁶; Sunderland-Groves, Jacqueline⁷; Sitoe, Almeida^8
1RSAC Ltd, UNITED KINGDOM; ²SarVision B.V., NETHERLANDS; ³Wageningen University, NETHERLANDS; ⁴Remote Sensing Solutions GmbH, GERMANY; ⁵TerraSphere Imaging & GIS B.V., NETHERLANDS; ⁶Equipe de Conservacao da Amazonia, BRAZIL; ⁷Borneo Orang-utan Survival Foundation, INDONESIA; ⁸Universidade Eduardo Mondlane, MOZAMBIQUE

The REDD Fast Logging Assessment & Monitoring Environment (REDD-FLAME) project has designed and demonstrated a system capable of monitoring tropical and sub-tropical forests using high-resolution, cloud-penetrating radar data (and optical imagery), acquired by Earth Observation satellites. By focussing on early detection of logging activities, the system provides the means to quickly identify the first signs of deforestation and thus contribute to the objectives of the UN-REDD (Reducing Emissions from Deforestation and Forest Degradation) Programme.

The system forms a high-resolution add-on for existing (semi-) operational low- to mid-resolution systems, providing hot-spot monitoring for areas at highest risk of deforestation, no matter what the weather. As such, it could be integrated into national or regional forest monitoring centres as an alert service facilitating timely intervention and thus reducing unauthorised tree loss in the context of carbon accounting.

The system has been developed with the support of the EC FP7-SPACE RTD Programme in collaboration with investigators in countries on three continents, in an effort to build lasting partnerships and share expertise. Test sites have been chosen to represent a variety of forest types and deforestation issues, and thus to prove the system’s versatility.

The main results of the project will be presented. REDD-FLAME is proven to be a reliable means to identify deforestation events quickly enough to allow intervention by the authorities and protection of undisturbed forest tracts. The system is ready to exploit Sentinel data as soon as this becomes available. Recommendations for operational implementation within national forest monitoring systems will also be discussed.

[For consideration for the special REDD Session]