Characterizing bare-earth elevations from airborne LiDAR data in a shrub-dominated mountain environment (Invited)

Airborne LiDAR (Light Detection and Ranging) data have made it possible to generate high resolution bare-earth digital elevation models (DEM). Characterizing bare-earth elevations derived from LiDAR data for areas with high variability in topography and vegetation, however, is a challenging task due in part to LiDAR sensor limitations and, the algorithms used to classify LiDAR returns, and the structural properties of dense shrubland vegetation. Most algorithms used for LiDAR return classification, or filtering, require prior specification of parameters before they can separate ground and non-ground returns. Using airborne LiDAR data from Reynolds Creek Experimental Watershed in southwestern Idaho, we present a method to automatically assign filtering parameters based on the lidar-derived characterization of the diverse vegetation and topography features that exist in this watershed. We have combined and enhanced the strength of existing height filtering algorithms to produce a hydrologically-valid DEM in a shrub dominated mountain environment. This new workflow for characterizing bare-earth elevation will be useful for monitoring and measuring snow processes, and other applications in hydrology, ecology, and geology.