Mapping Defoliation with Lidar

The aim of this article is to present a concept of using airborne laser scanning (LIDAR), with one scan only, to map defoliation as a forest health variable. The idea is to apply two independent algorithms on the LIDAR data set, to produce both actual and expected leaf area index (LAI) values for every cell in a grid over the area. LAI is estimated based on laser pulse penetration through the canopy layer, and expected LAI values are derived from stand density based on position and height of single trees as obtained from a single-tree segmentation algorithm. The results are preliminary findings from four ongoing and related studies. In the first study repeated laser scans had close to equal extinction coefficients for LAI estimation although the instruments and flight specifications were different. In the second study, based on the findings in the first we derived normal LAI values from extisting and large scale data sets with LIDAR and field data. The main independent variable was stand density, defined as the ratio between mean tree height and mean distance between the trees. The ratio between LAI and stand density was around 0.5, and this is a preliminary standard for a healthy pine forest. In a third study the woody area fraction of LAI was estimated from 14 total harvested trees, and turned out to be slightly below 50% for a healthy pine tree, which means that a totally defoliated pine forest would have an LAI/stand density ratio around 0.2. In the fourth study, these LAI standard values were confirmed with LIDAR data from a severe insect defoliation event in Norway 2005. In conclusion, the present preliminary results demonstrate a potential for application of airborne laser scanning for monitoring or mapping of defoliation as a forest health variable. * Corresponding author