3-D remote sensing of woody canopy height and carbon stocks by helicopter-borne scanning lidar

Abstract

A method for 3-D remote sensing of forest canopies using high-resolution, helicopter-borne scanning lidar is presented. The lidar device can scan almost all the ground surface with high resolution because a laser beam with a small footprint and a high scanning rate illuminates the ground surface from a slow-moving helicopter. The method permits the generation of 3-D images such as a Digital Elevation Model, Digital Terrain Model, and Digital Canopy Height Model (DCHM). The validity of the method was demonstrated in two applications. First, we estimated tree height from a DCHM of a forest on a steep slope, and found that errors were within 0.47 m for tree height (0.19 m RMSE). These results show that the estimation of tree height was greatly improved compared with estimates in previous studies. Second, we estimated carbon stocks in each tree and in the stand as a whole. From lidar-derived tree heights and allometric relationships between tree height and carbon stocks, we accurately estimated total carbon stocks of each tree in a coniferous Japanese cedar (Cryptomeria japonica) forest as well as carbon stocks at the stand scale.