Estimation of forest resource conditions in larix kaempferi stand combining high resolution digital aerial photography and DTM by LiDAR data

Abstract

The purpose of this research was to develop high resolution digital aerial photography to estimate fundamental stand parameters (number of standing trees, mean tree height, stem volume) in a Larix kaempferi stand. In order to identify tree species, a five-class (Japanese larch, Japanese red pine, broad-leaved trees, no vegetation, shadow) vegetation map was created with 90% accuracy from a multispectral digital orthophotograph. We extracted individual tree crowns by using a digital surface model (DSM) made from a stereo pair of panchromatic images and a digital terrain model (DTM) made from LiDAR data and, by overlaying the vegetation map, we were also able to determine the species of the crowns. Stand parameters were estimated for every tree crown. When we verified the accuracy of the stand parameter estimates with complete enumeration, we found that the accuracy rate for the number of standing trees was 85.6%, the difference in mean tree height was 0.4 m, and the accuracy rate for stem volume was 84.9%. Because of the difference in crown form, our estimates were more accurate for the conifers than for broad-leaved trees.