Determination of stand parameters by using UAV data

In accordance with the improvements in the accuracy of Unmanned Aerial Vehicles (UAV)-supported remote sensing data, their use in forestry studies has become widespread. In recent years, stand parameters can be calculated using UAV-based three-dimensional (3D) point cloud data. In this study, it was aimed to determine tree parameters (tree height, crown width and tree diameter) using UAV data in two sample areas with different vegetation understory density (low and high) in a stone pine (Pinus pinea, L.) stand. First, tree height and crown width parameters were obtained by using 3D point cloud generation and processing method with Structure for Motion (SfM) on UAV images taken from sample sites. To evaluate the success of the UAV-based method, height, crown width and diameter data of the trees in the study area were obtained using classical terrestrial measurement devices. Then, regression models that estimate the diameter of sample trees were developed using tree height and crown width data (independent variables) obtained from ground measurements. Finally, diameter values were calculated by entering the tree height and crown width values obtained by the UAV-based method into these models. In the study, the relationships between stand parameters determined by ground measurements and UAV data were examined with statistical analysis. In addition, the effectiveness of the UAV-based method was evaluated by comparing the stand parameters obtained with both methods with the Mean Absolute Error (MAE), Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) methods. According to the results, it was determined that there was a linear relationship (R2) between the tree height and crown width values determined by ground measurements and the values determined by UAV data. It was observed that the increase in understory density negatively affects the estimation of stand parameters of the method based on UAV data and therefore the relationship ratio was relatively lower. The results showed that the tree height and crown width parameters determined by UAV data generally had acceptable accuracy when compared to ground measurements. Error values (RMSE) for tree height were determined as 0.408 and 0.377 in the first and second sample area, respectively. For crown width, error values were found to be 0.446 and 0.497, respectively. When tree height and crown width parameters were compared, it was seen that tree height was determined with higher accuracy in both sample areas. When the sample areas were compared, it was determined that the crown width accuracy was higher in the sample area where the density of understory density was low.