Linking Carbon Fluxes With Remotely-Sensed Vegetation Indices For Leaf Area And Aboveground Biomass Through Footprint Climatology

Craig Wayson, USDA Forest Service, cwayson@fs.fed.us (Presenter)
David Hollinger, USDA Forest Service, davidh@christa.unh.edu
Nick Skowronski, USDA Forest Service, nskowronski@fs.fed.us
Richard Birdsey, USDA Forest Service, rbirdsey@fs.fed.us

To paramterize ecosystem models from carbon-flux data from eddy-flux towers efforts must be made to select data that best represents the region being modeled as well as linking the fluxes to remotely-sensed data that can be estimated from site to regional scales.

It is a major problem of bottom-up scaling that in-situ flux observations are in general spatially limited and disparate. Thus, to achieve valid regional exchange rates, models must be used to interpolate and extrapolate spatial domain covered by these observations. Observed and modeled fluxes can only be linked if they represent exchange over the same ecosystem. Because most long-term flux stations are not situated in spatially extensive homogeneous locations, this requirement is often a problem, but can be satisfied by selecting observation periods whose flux footprints are statistically representative of the type of ecosystem identified in the model. The flux footprint function indicates the time-varying surface “field-of-view” (or spatial sampling window) of an eddy-flux sensor, oriented mostly in upwind direction. For each observation period, the modeled flux footprint window is overlaid over a high resolution vegetation index map (derived from Landsat or a better resolution platform), to determine a footprint-weighted vegetation index for which the observation is representative.

Earlier work done using flux-footprint analysis to link fluxes to ecosystem models using just EVI showed a positive trend between EVI and eddy covariance measured fluxes, but the linkage was not strong. Leaf area is linked with C uptake, but forests tend to maximize leaf area early on with younger and older forests having similar leaf areas. Adding another remotely-sensed dataset (biomass) helps capture the processes of lower light use efficiency (as biomass increases per unit of leaf area there is a decline, due to the forest ageing) and the C losses due to respiration, both heterotrophic and autotrophic.

This work links carbon fluxes to remotely-sensed variables of enhanced vegegatation index (EVI) from Landsat TM imagery, biomass from LIDAR and a combined EVI-biomass layer to examine the representativeness of flux footprints to a larger region in the northern hardwoods in New Hampshire.

Presentation Type:  Poster

Session:  Coupled Processes at Land-Atmosphere-Ocean Interfaces   (Mon 4:00 PM)

Associated Project(s): 

• Related Activity

Poster Location ID: 91