A New Method for Incorporating Hillslope Effects to Improve Canopy-Height Estimates From Large-Footprint LIDAR Waveforms

Forest structure variables, such as the canopy height, are of central interest for the quantification of ecosystem functions and the assessment of biomass levels. The objective of this letter is to propose a new method for ridding canopy-height estimates from the influence of the hillslope within large-footprint (light detection and ranging) LIDAR waveforms. The method is based on modeling (using two generalized Gaussian functions) and the fitting of canopy and ground components to large-footprint (30 m) waveforms. The canopy heights were estimated for 27 waveforms: A root-mean-square error of 3.3 m was obtained using a high-resolution digital terrain model (DTM) to estimate the ground component (4.3 m using the 80-m-resolution Shuttle Radar Topography Mission DTM) and 3.5 m when self-estimating the ground component (hillslope) based on the large-footprint waveform. This approach opens new possibilities for waveform decomposition for natural resources and topography assessments based on large-footprint LIDAR waveforms in forest environments.