Ricciuto, D.M., M.P. Butler, K.J. Davis, B.D. Cook, P.S. Bakwin, A.E. Andrews, and R.M. Teclaw, submitted. A Bayesian synthesis inversion of simple respiration and GEP models with eddy covariance data in a northern Wisconsin forest: Determining the causes of interannual variability, Agricultural and Forest Meteorology.

Variability in fluxes of carbon dioxide for the years 1997-2001 was analyzed from the WLEF tall tower in northern Wisconsin. Estimates of uncertainty in sums of NEE due to turbulent variability and gap-filling were calculated. Although absolute values of NEE sums were affected by the choice of a friction velocity (u*) threshold, this choice did not affect estimates of interannual variability. WLEF was also observed to be a source of carbon to the atmosphere regardless of the u* threshold applied. Estimates of uncertainty were generally smaller than the observed range of seasonal and interannual variability in tower fluxes. Daytime fluxes were observed to be more variable than nighttime fluxes, and spring and summer fluxes were observed to be more variable than autumn and winter fluxes. This variability is strongly correlated with changes in climate variables. Soil moisture and soil temperature were found to be the primary drivers of interannual variability in seasonal sums of NEE. The standard nonlinear gap-filling regression models of ecosystem respiration and gross ecosystem productivity were extended to incorporate these long-term climate effects, and these models were calibrated using Bayesian inversion techniques. In most cases, these models produced statistically significant correlations with interannual variability of observed season-averaged values of ecosystem respiration and of estimated season-averaged values of gross ecosystem productivity.