engleski

Understorey identification through the generation of Canopy Base Height Models based on LiDAR data

The aim of this work is to find feasible methodologies to identify the canopy of understorey below a higher forest canopy using the information on vertical structure of vegetation that LiDAR point clouds provide. Canopy Base Height (CBH) identification could be the first step to extract understorey formations from LiDAR returns. The study develops ad hoc CBH models as a previous phase to set the optimum heightbreak between LiDAR returns coming from tree canopy and from understorey. Canopy base height models have been built at individual-tree level, with tree size, competition indexes and stand variables as predictors. In the subsequent step, we have used two alternatives to obtain the understorey height model: 1) Tree-level approach, using each tree modelled CBH to derive return heightbreak, and 2) Plot-level approach, where a wall-to-wall model of the mean CBH in each cell is used to establish the optimum heightbreak. Both methodologies have been performed on different LiDAR densities to check the precision of the models under diverse types of LiDAR data and to know which density of LiDAR returns is the optimal to build these models. Our research shows that utilising CBH models can be useful to establish an optimum threshold to separate tree canopy and underbrush lidar returns. This methodology is able to separate the different formations in an irregular forest through their vertical distribution. Tree-level results adjust better to the real stand conditions and offer a more realistic distinction between the returns that belong to the canopy and those that belong to the understory vegetation.

understorey vegetation ; Airborne Laser Scanning (ALS) ; LiDAR ; Canopy Base Height Models

nije evidentirano