Large-area stem volume estimation based on airborne laser scanner data and national forest inventory data

Wien

2009

5 S

Monographie

10569S

161222

The objective of this paper is to evaluate a new approach for the estimation of stem volume based on ALS data. The approach was tested for the entire Vorarlberg province in Austria, which covers about 97,000 ha of forests. ALS data were acquired in the framework of a commercial district-wide terrain mapping project and were acquired during several winter- and summer flight-campaigns. The laset point density varies between 1 and 4 points/m2. The pre-processing of the ALS data was performed with state-of-the-art methods and commercially available software packages. A novel linear model was employed, which describes the stem volume as a linear function of canopy volume. The canopy volume is defined by the volume between the terrain surface and the canopy surface derived from the first-echo hits. The calibration of the stem volume model was based on national forest inventory (NFI) data. The first results of this approach provided a maximum R2 of 0.69 and a standard deviation (SD) of residuals derived from cross-validation of 129.9 m3/ha (34.6%). A stratification of the forest into coniferous, mixed and deciduous forests increased the achieved accuracies (e.g. R2 = 0.78; SD = 110.9 m3/ha (29.0%) for coniferous forests). As all involved data sources are in operational use, it is ensured that the results obtained in this study are relevant for forest and environment policy and practical applications.