Estimating tropical-forest density profiles from multibaseline interferometric SAR

Abstract

Vertical profiles of forest density are potentially robust indicators of forest biomass, fire susceptibility and ecosystem function. Tropical forests, which are among the most dense and complicated targets for remote sensing, contain about 45% of the world's biomass. Remote sensing of tropical forest structure is therefore an important component to global biomass and carbon monitoring. This paper shows preliminary results of a multibasline interfereomtric SAR (InSAR) experiment over primary, secondary, and selectively logged forests at La Selva Biological Station in Costa Rica. The profile shown results from inverse Fourier transforming 8 of the 18 baselines acquired. A profile is shown compared to lidar and field measurements. Results are highly preliminary and for qualitative assessment only. Parameter estimation will eventually replace Fourier inversion as the means to producing profiles.