IMPROVING LiDAR-BASED FOREST RESOURCES ASSESSMENT USING PRECISE FIELD VALIDATION TECHNIQUES OF TREE INVENTORY DATA

ABSTRACT: Tree inventory is a necessary tool to validate the accuracy of acquired LiDAR point cloud against the gathered field data when extracting information on forest resources. This research study focuses on establishing a protocol for field inventory survey and comparison of positional accuracies of surveyed points from different methods. A 2-ha permanent forest plot in Molawin-Dampalit, Mount Makiling, Laguna, established by the Makiling Center for Mountain Ecosystems (MCME) using a hand-held GPS and bearing method, was used as the comparative study site for this research. Diameter at breast height of trees with diameters greater than 10 cm, 1-10cm and less than 1 cm were measured using a diameter tape at 20mx20m, 10mx10m and 5mx5m grids,respectively. Canopy closure and gap fraction was measured using digital hemispherical photography and fisheye lens. Regression methods were used to establish relationships and to know if the parameters directly measured from LiDAR specifically canopy height, point density, point spacing and canopy cover would derive the diameter at breast height of the trees in the study site. Based on the field inventory, tree height in the study site ranges from 0.01 to 33m, DBH ranges from 0.01 to 150cm, and canopy cover ranges from 53.5 % to 98.3%. Total Station survey further showed higher positional accuracies of the surveyed trees. Consequent field daa cases studied also showed that field and LiDAR-derived DBH becomes comparable at 10x10 grid and pixel size. 1