You are here

Harvard Forest >

Harvard Forest Research Project 2022

  • Title: Linking Leaf Traits and Canopy Structure to Remotely Sensed Vegetation Indices
  • Principal investigator: Mark Friedl (
  • Institution: Boston University
  • Primary contact: Mark Friedl (
  • Team members: Mark Friedl; Leticia Lee; Minkyu Moon
  • Abstract:

    This research complements and extends previous efforts designed to understand functional relationships among forest canopy function, leaf traits and phenology. In previous activities, we used a combination of ground-based measurements, PhenoCam data, meteorological observations, and time series of remotely sensed imagery to develop models of forest phenology. The goals of these activities were (1) to improve understanding of how remotely sensed observations can be used to study and model the green leaf phenology of mixed temperate forests from local to regional scale, and (2) to improve understanding of how climate change is impacting the phenology of forests. In this research we will extend these efforts by collecting data that will allow us to empirically link the phenology of leaf and canopy traits to time series of remotely sensed imagery. To accomplish this, we will collect leaf level photosynthesis measurements, high-spectral resolution leaf reflectance spectra, and nitrogen content for leaves located in the upper canopy of multiple trees for each of the dominant deciduous species at Harvard Forest -- Red Oak (Quercus rubra), Red Maple (Acer rubrum), and Yellow Birch (Betula alleghaniensis). Following the general approach used by Dillen et al. (2012), measurements will be collected for several trees at different points in the growing season, including fall senescence, using the Harvard Forest canopy access lift. Information derived from previously acquired unmanned aerial vehicle imagery will be used in association with LiDAR imagery acquired by the NEON airborne observatory platform (AOP) will be used to develop a first order representation of 3-dimensional canopy structure. High spatial resolution hyperspectral information from the NEON AOP will be used to map species composition, and Landsat and Sentinel 2 will provide time series of multispectral data over the growing season. Together, these data sets will be used to develop a first-order model that relates space-borne multispectral measurements to leaf traits, species composition, and canopy structure. In the long run, the goal is to use the general framework developed through this effort to improve models and remote sensing-based methods for estimating seasonal scale changes and dynamics in forest light use efficiency.