Improved Method to Calculate Urban Forest Vertical Structure Using Airborne Laser Scanning Data

Understanding the role of urban forests requires high-resolution information for characterizing their vertical structure. Light Detection and Ranging (LiDAR) point clouds with a density of at least 8 points/m2 were classified on the example of Zęborzyce forest in Lublin (Poland). A new method for analysis of the urban forest vertical structure was implemented using a division of a cloud of points into cells with a description of X, Y, and Z coordinates and density with further calculation of the Pearson index. Using the profile approximation, we obtained a reference distribution of point density by height. A classification accuracy of 95% for the urban forest has been achieved. The proposed method for analyzing the vertical structure of urban forests enables fast and realistic modelling for the forests of the entire city. The usefulness of LiDAR data and statistical analyses for estimating the vertical structure of urban forests was shown. The final classification was accomplished using these pre-classification reference maps accepted as ground truths. Accuracy assessment of the resulting map showed good performance, with an overall accuracy of 88%. These results exhibit the potential of LiDAR data application for estimating urban forest vertical structure. The proposed method contributes to the efficient and effective management of urban forests.