Article ID Journal Published Year Pages File Type
82601 Agricultural and Forest Meteorology 2008 11 Pages PDF
Abstract

The three-dimensional (3D) distribution of plant area density (PAD) is difficult and time-consuming to determine. In this study, we measured this distribution using a ground-based laser scanner consisting of a laser range finder and two rotation stages, and compared the results with those obtained based on the attenuation of solar radiation and litter traps. To determine the 3D distribution of PAD, we divided the forest horizontally and vertically to define volume elements (voxels). In each voxel, we calculated PAD from the gap fraction, which was defined as the probability that laser beams from the forest floor would pass through the voxel without interception. Measurements were obtained four or five times between the in-leaf and leafless periods in 45- and 46-year-old Japanese larch (Larix kaempferi Sarg.) plantations with different stem densities after thinning. We found the following: (1) Because the largest tree height was greater than the highest point of interception of the laser beams, our method underestimated PAD in the upper layer of the forest. (2) Compared with estimates based on the attenuation of solar radiation, our method overestimated plant area index (PAI) by around three units throughout the season, but successfully captured the reduction of PAI between bud break and leaf-fall. (3) The horizontal distribution of PAI estimated by our method was well correlated with that estimated using litter traps and reproduced the horizontal non-uniformity of PAI. Despite some limitations, such as the diameter and divergence angle of the laser beam, the ground-based laser scanner successfully provided direct, non-destructive estimates of the forest's 3D distribution of PAD.

Related Topics
Physical Sciences and Engineering Earth and Planetary Sciences Atmospheric Science
Authors
, , , , ,