Spatial patterns of leaf angle distribution covary with canopy fluorescence yield, reflectance indices, and leaf chlorophyll content, in a mixed temperate forest

Plant canopies are integrated units that coordinate their functional (e.g., foliar biochemistry) and structural properties. This coordination affects remote sensing observations of canopy reflectance and solar-induced chlorophyll fluorescence (SIF). One key canopy structural property is leaf angle. Despite the fact that radiative transfer models have shown the crucial role of leaf angle in modulating remote sensing signals, methodological and technological barriers have prevented detailed investigations of how leaf angle covaries with canopy function and remote sensing observations. In this study, we employ a novel uncrewed aerial system (UAS) called FluoSpecAir to study the spatial patterns in far-red (FR) SIF (SIF_obs,FR), near-infrared reflectance and radiance of vegetation (NIR_V and NIR_VR), normalized difference vegetation index (NDVI), and chlorophyll:carotenoid index (CCI), across individual tree canopies during two separate time periods. Additionally, we collected 3D scans of individual tree canopies using terrestrial laser scanning (TLS) and estimated foliar pigment content from leaf reflectance spectra. We used the 3D scans to calculate the leaf angle distribution (LAD) and leaf area voxel density (LAVD) of each canopy. We modeled LAD using a beta distribution, which is parameterized by μ and ν, and the leaf inclination distribution function (LIDF), which is parameterized by LIDFa and LIDFb. We found that ν and μ, which are inversely related to the variance in leaf angle, covaried with spatial patterns in peak growing season canopy CCI, NDVI, SIF_obs,FR, and $\frac{{\mathrm{SIF}}_{\mathrm{obs},\mathrm{FR}}}{{\mathrm{NIR}}_{V}R}$, and leaf chlorophyll content. Canopies with greater variation in LAD, thus lower ν and μ, have larger values of NDVI, CCI, SIF_obs,FR, $\frac{{\mathrm{SIF}}_{\mathrm{obs},\mathrm{FR}}}{{\mathrm{NIR}}_{V}R}$, and leaf chlorophyll content, while LAVD is not correlated with these remote sensing metrics. We found positive correlations between leaf chlorophyll content and canopy NDVI, SIF_obs,FR, and $\frac{{\mathrm{SIF}}_{\mathrm{obs},\mathrm{FR}}}{{\mathrm{NIR}}_{V}R}$, as well. Together, our results show that across our study site during the peak growing season, spatial variability in remote sensing variables is driven by the coordination between LAD and leaf chlorophyll content. These findings provide important context for how we interpret landscape level variability in SIF and $\frac{{\mathrm{SIF}}_{\mathrm{obs},\mathrm{FR}}}{{\mathrm{NIR}}_{V}R}$, and how spatial variation in both can be used to infer differences in plant metabolism.