Model-Based Estimation of Forest Canopy Height and Biomass in the Canadian Boreal Forest Using Radar, LiDAR, and Optical Remote Sensing

One of the fundamental technical challenges of any new spaceborne vegetation remote sensing mission is the determination of what sensor(s) to place onboard and what, if any, overlapping modes of operation they will employ as each onboard sensor adds significant cost to the overall mission. In this article, the remote sensing of forest parameters using multimodal remote sensing is presented. In particular, polarimetric radar, Light Detection And Ranging (LiDAR), and near-IR passive optical sensing platforms are employed in conjunction with physics-based models. These models are used to accurately estimate forest aboveground biomass as well as canopy height in homogeneous areas. It is shown that this proposed method is capable of achieving high accuracy estimates while using minimal ancillary data in the estimation process. We present a method to combine measured data sets with our geometric and electromagnetic sensor models to develop a forest parameter estimation algorithm that fuses multimodal remote sensing technologies with a minimal amount of ground information and yields an accurate estimate of forest structure including dry biomass and canopy height with rms errors of 1.6 kg/m(2) and 1.68 m respectively.