An assessment of the utility of an unmanned aerial vehicle for measuring tree heights in a jack pine provenance trial

Abstract

This independent study is an assessment of the utility of an unmanned aerial vehicle (UAV) for measuring the heights of jack pine (Pinus banksiana) trees in a provenance trial. An electronic hypsometer, the Haglöf Vertex IV, was used in the field to measure tree heights, and an UAV, the DJI Mavic Pro 1, was used to collect a point cloud of the provenance trial. Agisoft PhotoScan was the cloud-processing engine used for image processing and ArcMap was used for estimating tree heights. Tree height measurements were recorded for 256 jack pine in the study area. Measurements taken using the Haglöf Vertex IV displayed greater heights in comparison to the heights derived from the UAV imagery. The mean difference between the two datasets was 1.1 m with a standard deviation of 0.8 m. A correlation between the two datasets was observed when mean tree height per provenance was compared; a value of 0.9 was calculated using the Pearson Product-Moment Correlation Coefficient. This value of 0.9 indicated a positive association between the two datasets, however, ranking provenance means from highest to lowest indicated the two datasets differed. Although the UAV was capable of measuring tree heights, this study demonstrates that further understanding of measurement error is important for producing accurate results.