Exploring horizontal area-based metrics to discriminate the spatial pattern of trees and need for first thinning using airborne laser scanning

Abstract

Summary The objectives of this study were (1) to identify point cloud metrics and horizontal texture and landscape metrics from airborne laser scanning (ALS) data, which can be used to determine the spatial pattern of trees and, further, the need for first thinning and (2) to study if the clustered spatial pattern of trees and the need for first thinning can be separated from other spatial patterns and need for thinning classes. The field data consisted of 28 microstands, which had reached the first thinning height but had not yet been thinned. Linear discriminant analysis was used to classify microstands by means of the metrics calculated from the ALS data. A classification based on the spatial pattern of trees discriminated stands with the overall accuracy (OA) being 0.89 and kappa-value (k) 0.77. Similarly, a classification based on the need for first thinning was also successful (OA = 0.96 and k = 0.93). Horizontal landscape metrics were found to be good predictors of the spatial patterns of trees, whereas the landscape and point cloud metrics were found good predictors of the need for first thinning.