Biomass estimation in a tropical wet forest using Fourier transforms of profiles from lidar or interferometric SAR

Abstract

Tropical forest biomass estimation based on the structure of the canopy is a burgeoning and crucial remote sensing capability for balancing terrestrial carbon budgets. This paper introduces a new approach to structural biomass estimation based on the Fourier transform of vertical profiles from lidar or interferometric SAR (InSAR). Airborne and field data were used from 28 tropical wet forest stands at La Selva Biological Station, Costa Rica, with average biomass of 229 Mg‐ha −1 . RMS scatters of remote sensing biomass estimates about field measurements were 58.3 Mg‐ha −1 , 21%, and 76.1 Mg‐ha −1 , 26%, for lidar and InSAR, respectively. Using mean forest height, the RMS scatter was 97 Mg‐ha −1 , ≈34% for both lidar and InSAR. The confidence that Fourier transforms are a significant improvement over height was >99% for lidar and ≈90% for InSAR. Lidar Fourier transforms determined the useful range of vertical wavelengths to be 14 m to 100 m.