Is flux divergence in the tower layer important in estimating annual NEE using eddy-covariance measurements?

Carbon and energy fluxes measured at the University of Michigan Biological Station (UMBS) AmeriFlux site in the first three years (1999-2001) were published recently. Estimated annual NEE based on long-term eddy-covariance measurements are sensitive to criteria used for data quality control, including the friction velocity threshold, as well as gap-filling methods. It was reported that annual NEE estimate based on eddy-covariance measurements at a higher observation level (~2.1hc, where hc = 22 m is the mean canopy height) is consistently smaller (by 23% to 53%) in absolute magnitude than that observed at a lower level (1.5hc) over each of the three years, in part due to apparent greater CO2 fluxes observed at the higher level during the dormant season. In the current study, we first present variations of flux divergence in different stability conditions and seasonal changes of canopy morphology for CO2, water vapor, sensible heat and momentum. A comparison will also be made to other issues in estimating carbon, water and energy fluxes using eddy-covariance method, such as averaging time length, choice of coordinate rotation, cospectral correction, etc. Eddy-covariance measurements at two heights (1.8hc and 1.3hc, where hc = 26 m) at the Morgan-Monroe State Forest AmeriFlux site may also be discussed.