Unlocking the Power of Active Remote Sensing for Ecosystem Services Modeling in the Amazon’s Forest-Agriculture Interface

The last decade has experienced a dramatic improvement in the extent and consistency of tree cover and gross deforestation products from optical imagery. These optical-based datasets are capable of tracing forest clearings made for plantations and pastures, but they suffer from cloud cover and may lump mature forest, secondary forest, and plantations into a single ‘forest’ class. This project augments existing deforestation monitoring systems by filling two important knowledge gaps: (1) characterization of landscapes beyond binary forest / non-forest classification – required in order to strategize zoning, monitoring, and enforcement; and (2) identification of the commodities that potentially replace forest stand once deforestation is detected. Characterizing these production systems is crucial for decision makers who manage land tenure systems, improve monitoring, and design conservation strategies within the context of global commodity markets. The project will: (1) expand land cover maps by identifying oil palm and cacao plantations and secondary forests; (2) develop a prototype tool that utilizes plantation maps to assess compliance by farmers enrolled in Zero Deforestation Supply Chain initiatives; and (3) transfer knowledge and technology to end-users while addressing locally-relevant questions.