Imaging spectroscopy is increasing in prominence to estimate forest functional traits and characterize ecosystem patterns and processes across large spatial scales. The standard approach entails statistical models that relate spectral signatures to foliar properties, resulting in spatially comprehensive maps of forest traits. These models validate well, but the influence of canopy structure on trait retrievals is poorly understood. Canopy structure is known to influence remote sensing measurements through a combination of shading, gaps, bidirectional reflectance and multiple scattering. Here we investigate the effects of canopy structure on foliar trait estimates derived from Airborne Visible InfraRed Imaging Spectrometer (AVIRIS) imagery in a mixed temperate forest in northern Wisconsin. Leaf-on LiDAR data were used to estimate canopy structural properties including canopy closure and height and were compared against key foliar traits including nitrogen, carbon and lignin.
Results/Conclusions
The results indicate that structural effects are present in varying degrees, depending in part on species present, but that image processing methods can reduce the impacts of structure on retrieval algorithms.