Abstract The northern hardwood forests of Eastern Canada, particularly stands dominated by sugar maple (Acer saccharum Marsh.), are facing ongoing decline due to historical and contemporary environmental pressures. Traditional ground-based crown assessments for tree health are essential for management, but can be subjective, costly, and limited by their viewing perspective. While conventional remote sensing methods can effectively capture tree crown structure from above, and terrestrial approaches can capture the stems and crowns from beneath, occlusion by the dense crowns of mature sugar maples makes capturing reliable estimates challenging. We examine the potential of intra-canopy aerial drone-based photogrammetry, involving flights beneath, within, and above tree crowns, to generate detailed 3D point clouds of 29 sugar maple trees in Quebec, Canada. From these point clouds, we derived estimates of key structural attributes including diameter at breast height (DBH), tree height, and crown base height (CBH). We used ray-marching to quantify crown transparency across 162 viewing angles, forming a sphere around the crown, and compared predictions to ground-based visual estimates and health categories. Photogrammetric estimates had significant correlations with ground-measured attributes, including DBH (r = 0.82), tree height (r = 0.55), and CBH (r = 0.73 and 0.78 across two distinct definitions). Modeled crown transparency correlation was also significant when compared to ground-based visual assessments (ρ = 0.54), suggesting that intra-canopy drone-based photogrammetry can offer rapid and objective assessment of crown condition.
Olson et al. (Mon,) studied this question.