ZLV Online

Abstract

Area-based approach of forest inventory is the statistical method, where the features and predictors are assessed from the laser derived surface models and point clouds, which are directly used for forest parameters estimation, typically using linear regression or k-nearest neighbor imputation method. Collecting of field data for the creation of model for forest stand volume estimation was carried out at the Training Forest Enterprise, Křtiny (Czech Republic). In total, 43 plots with a radius of 12.62 m were selected, and diameters at breast-height and heights of all trees were measured on each of them. Selected LiDAR metrics calculated in FUSION software were used as independent variables for the creation of linear regression equations. Multivariate linear regression models were created using the software QC Expert for all forest variables. Models for volume and basal area estimation explained 82% or 72% of all cases. Regression triplet was tested for the final regression models. Comparison with the Forest Management Plan was made using zonal statistics applied to maps of mensurational variables. Deviation depended on the complexity of a particular forest stand structure.