Constraining uncertainty in future model projections of forest dynamics requires adequately capturing biogeographic variation in carbon storage and turnover. Incorporation of edaphic controls on this variation is key, particularly in the tropics. We developed an approach for integrating airborne remote sensing data with a terrestrial biosphere model to constrain heterogeneity in carbon flux dynamics across communities that span topographic-related edaphic gradients, using data from two tropical forests in Malaysian Borneo. Using leaf and plant traits related to structure (height, gap fraction, LAI, and canopy profiles), defense (lignin and phenols), and productivity (SLA, foliar N, and foliar P), derived from Visible-Shortwave Imaging Spectrometer (VSWIR) measurements collected by the Global Airborne Observatory, we characterized community-scale differences in function to define initial conditions for a terrestrial biosphere model (ED2). We then parameterized plant functional types based on remotely sensed community trait distributions (SLA, Vcmax, and gap dynamics). Using a simple PCA and cluster analysis, we find that remotely sensed traits alone can be used to distinguish distinct functional communities, explaining nearly 80% of variance across these communities. It also revealed important structural and functional variation within an area characterized as a single forest community based upon field measurements. Plant functional types parameterized with site constrained trait and disturbance values yielded more accurate predictions of canopy demography, forest productivity and above-ground biomass dynamics. Since the main axes of variation in leaf traits correspond to quantities that are measurable from the planned spaceborne imaging spectroscopy missions such as NASA’s Surface Biology Geology satellite imaging spectroscopy mission, our approach offers a framework for model-data integration that can be tested and employed across the tropical forest ecosystems at regional and potentially global scales.
Ecosystem demography model, imaging spectroscopy, LiDAR, Borneo, Malaysia, functional ecology