Application Of Unmanned Aerial Vehicle (UAV) Remote Sensing Technology for Estimation of Tree Height in Heterogeneous Forest

Tree height is an important piece of information in forest management. Cost, time, and effort are some of the limiting factors in extracting tree height values on a large scale. The canopy height model approach through aerial photography using UAV can be used to quickly estimate the height of large-scale trees combined with field measurements. CHM analysis was carried out using spatial statistics to get the maximum tree height value based on the tree canopy. Evaluation of accuracy in the form of statistical tests is used to assess the level of accuracy of the estimation. Photogrammetry results show that the obtained CHM has a resolution of 11.8 cm/pixel with the results of the evaluation of tree height accuracy having an RMSE of 2.4 m, MAE 2.0 m, SDE 3.8 m. The chi-square statistical test shows that the results of the tree height estimation accept H0 and there is a strong relationship between the observed tree height and the estimation through linear regression with an R2 value of 0.67. The broad estimation of height shows that Mansinam Island has a tree height in the range of 7 – 66 m. The dominant tree height is in the 19-30 m class with the number of individuals reaching 1,877 trees. This study shows that CHM obtained from aerial photography using low-cost UAVs is still able to estimate tree height well. For future studies, it is necessary to use a ground control point (GCP) to increase the accuracy of the elevation model and orthophoto.