Scalar Profile Assimilation Into a Multi-Layer Model of Canopy-Atmosphere Exchange: Toward Optimal Estimation of Net Ecosystem Exchange

Darren Drewry

Department of Civil and Environmental Engineering
Duke University

A major focus of current research efforts in land-atmosphere exchange is the modeling and quantification of carbon dioxide, water vapor and energy fluxes between forested ecosystems and the atmosphere. Mechanistic models of the coupled physical and biological processes that determine the magnitude of scalar fluxes have been developed and tested at many sites under a variety of environmental conditions. High frequency eddy covariance measurements of scalar fluxes are often used to test these canopy exchange models. Concurrent observations of carbon dioxide concentration profiles within the canopy airspace are frequently left unutilized in such modeling studies. We explore the assimilation of the information contained in these concentration profile measurements to constrain forward model estimates of net ecosystem exchange (NEE). A high-resolution, one-dimensional multi-layer model of canopy-atmosphere dynamics, including turbulent transport, vertical radiation interception, photosynthesis, stomatal conductance and respiration, is implemented and tested against eddy covariance flux measurements taken at the Duke Forest. A simple assimilation scheme is used to compare the results of forward model integrations both with and without the assimilated profile information. Implications are discussed for the optimal merger of data and models for the estimation of NEE.

Last Update: March 30, 2005