Modeling Semiarid Low-Height Vegetation with Airborne LiDAR

Low-height vegetation, common in semiarid regions, is difficult to characterize with airborne LiDAR (light detection and ranging) due to the similarities, in time and space, of the point returns of vegetation and ground. Other complications may occur due to the lowheight vegetation structural characteristics and the effects of terrain slope. This research is focused on modeling methods and error assessment of low-height vegetation in varying terrain. Several methods to best determine vegetation height and 2-d crown area are developed using both the LiDAR point cloud and rasters derived from the point cloud. These methods are tested on varying sloped terrain. Error assessments of bare earth terrain models in low-height vegetation cover types and slopes are also performed. Recommendations for modeling low-height vegetation and/or filtering lowheight vegetation from terrain models will be presented, along with open-source algorithms.