A LiDAR-based approach for a multi-purpose characterization of Alpine forests: an Italian case study

Several studies have verified the suitability of LiDAR for the estimation of forest metrics over large areas. In the present study we used LiDAR as support for the characterization of structure, volume, biomass and naturalistic value in mixed-coniferous forests of the Alpine region. Stem density, height and structure in the test plots were derived using a mathematical morphology function applied directly on the LiDAR point cloud. From these data, digital maps describing the horizontal and vertical forest structure were derived. Volume and biomass were then computed using regression models. A strong agreement (accuracy of the map = 97%, Kappa Cohen = 94%) between LiDAR land cover map (i.e., bare soil, forest, shrubs) and ground data was found, while a moderate agreement between coniferous/broadleaf map derived from LiDAR data and ground surveys was detected (accuracy = 73%, Kappa Cohen = 60%). An analysis of the forest structure map derived from LiDAR data revealed a prevalence of even-age stands (66%) in comparison to the multilayered and uneven-aged forests (20%). In particular, the even-age stands, whether adult or mature, were overwhelming (33%). A moderate agreement was then detected between this map and ground data (accuracy = 68%, Kappa Cohen = 58%). Moreover, strong correlations between LiDAR-estimated and ground-measured volume and aboveground carbon stocks were detected. Related observations also showed that stem density can be rightly estimated for adult and mature forests, but not for younger categories, because of the low LiDAR posting density (2.8 points m(-2)). Regarding environmental issues, this study allowed us to discriminate the different contribution of LiDAR-derived forest structure to biodiversity and ecological stability. In fact, a significant difference in floristic diversity indexes (species richness - R, Shannon index - H') was found among structural classes, particularly between pole wood (R=15 and H'=2.8; P <0.01) and multilayer forest (R=31 and H'=3.4) or thicket (R=28 and H'=3.4) where both indexes reached their maximum values.