Assessing the Impact of Central Appalachian Tree Species on Canopy Albedo via Measurement of Leaf Angles from Repeated Ground-based, Drone, and Hemispherical Photography

Abstract:

Satellite measurements of forest albedo are directly used in climate models, and could be used in models of the C and N cycles if we more fully understood the mechanism causing a strong correlation of forest albedo with canopy N and C assimilation. One attractive mechanism posits that tree species have evolved convergent leaf and canopy traits. While the leaf traits of tree species are known to drive variability in canopy N and C assimilation, linking tree species to variability in albedo is challenging because of the difficulty in measuring important canopy traits like leaf angle. To refine techniques for measuring leaf angle, and test the hypothesis that high albedo in the central Appalachians could be linked to the abundance of species with canopy traits of more horizontal leaf angles, we conducted four tests with ground-based, drone, and hemispherical photographs. First, we used a leveled camera on a steep slope to repeatedly, and directly measure the leaf angle of over 400 leaves within the canopies of oak, maple, and beech trees. Across all 21 repetitions (3 times a day on 7 dates between May and July), we observed consistent species differences in mean leaf angle (MLA), with maple always being the most horizontal (MLA = 14-18°) and oak the most vertical (MLA = 19-28°). Second, we again found highly significant species differences in MLA when we used a hexacopter drone with a camera on a self-leveling gimbal to make over 1020 direct measurements of leaf angle from six tree species in three broadleaf deciduous forest plots. Third, to measure MLA of a whole multi-species canopy, we compared a species abundance-weighted plot average of the drone-measured MLA values with an indirect, ground-based hemispherical photograph method. The strong agreement of these direct and indirect plot-level methods finally led us to compare a broader set of 61 plot-level hemispherical photo MLA measurements with canopy albedo measured by AVIRIS in broadleaf deciduous forests. In this last test, we found strong support (R² = 0.42, p < 0.0001) for the hypothesis that more horizontal MLA causes higher albedo. This final result, coupled with the observed strong species differences in MLA, lends support to the idea that the high albedo in the Central Appalachians may be linked to high abundances of tree species with canopy traits of horizontal MLA.