Linking the near- surface camera-based phenological metrics with leaf chemical and spectroscopic properties

Plant phenology is an important indicator of climate change. Near-surface cameras provide a way to continuously monitor plant canopy development at the scale of several hundred meters, which is rarely feasible by either traditional phenological monitoring methods or remote sensing. Thus, digital cameras are being deployed in national networks such as the National Ecological Observatory Network (NEON) and PhenoCam. However, it is unclear how the camera-based phenological metrics explain leaf physiology by linking with leaf chemical and spectroscopic metrics throughout the growing season.

We used the temporal trajectories of leaf chemical properties (chlorophyll a and b, carotenoids, leaf water content, and leaf carbon/nitrogen content) and leaf reflectance/transmittance to understand the temporal changes of camera-based phenological metrics (e.g., relative greenness), which was acquired from our Standalone Phenological Observation System installed on the island of Martha’s Vineyard and Harvard Forest, MA (dominant species: Quercus alba). Leaf chemical and spectroscopic properties were measured weekly from June to November, 2011. We found that the chlorophyll concentration showed similar temporal trajectories to the relative greenness. However, the chlorophyll concentration lagged behind the change of relative greenness for about 20 days both in the spring and the fall. The relative redness is a better indicator of leaf senescence in the fall than the relative greenness. We derived relative greenness from leaf spectroscopy and found that the relative greenness from camera matched well with that from the spectroscopy in the mid-summer, but this relationship faded as leaves start to fall.

This work suggests that we should be cautious to interpret camera-based phenological metrics, and the relative redness could potentially be a useful indicator of fall senescence.