Estimation of Tropical Forest Height and Aboveground Biomass from Dual-band InSAR measurements in Peruvian Amazon

In July 2009, Earthdata Inc. acquired Synthetic Aperture Radar (SAR) data over a lager part of the Peruvian lowland Amazon and mountain forests (more than 5000 km 2 ). The project was designed to provide high spatial resolution imagery to the science community to estimate and map forest aboveground biomass and to assess the capability of the measurements for REDD baseline applications. GeoSAR collects X-Band (VV, 9.7GHz), and P-Band, (HH, 0.35GHZ) interferometric synthetic aperture radar (SAR) data in single-passes, over swaths of width ~11km, that may extend to several hundred kilometres, and from which are derived digital elevation models (DEMs) [1]. In this study, we develop a fusion approach by combining the dual-band InSAR phase and coherence with backscatter power at P-band HH and HV polarizations to estimate forest vertical structure (height) and aboveground biomass. The results are verified using forest structure and estimated biomass over a suite of field inventory plots distributed across landscapes over the study area. The estimation errors and spatial uncertainties in forest biomass are examined to provide realistic guidelines for REDD baseline carbon stock assessments. 2. Methodology