To assess annual budgets of CO2 exchange between the biosphere and atmosphere over representative ecosystems, long-term measurements must be made over ecosystems that do not exist on ideal terrain. How to interpret eddy covariance measurements correctly remains a major task. At present, net ecosystem CO2 exchange is assessed, by members of the micrometeorological community, as the sum of eddy covariance measurements and the storage of CO2 in the underlying air. This approach, however, seems unsatisfactory as numerous investigators are reporting that it may be causing nocturnal respiration flux densities to be underestimated. A new theory was recently published by Lee (1998, Agricultural and Forest Meteorology 91: 39-50) for assessing net ecosystem-atmosphere CO2 exchange (N-e) over non-ideal terrain. It includes a vertical advection term. We apply this equation over a temperate broadleaved forest growing in undulating terrain. Inclusion of the vertical advection term yields hourly, daily and annual sums of net ecosystem CO2 exchange that are more ecologically correct during the growing season. During the winter dormant period, on the other hand, corrected CO2 flux density measurements of an actively respiring forest were near zero. This observation is unrealistic compared to chamber measurements and model calculations. Only during midday, when the atmosphere is well-mixed, do measurements of N-e match estimates based on model calculations and chamber measurements. On an annual basis, sums of N-e without the advection correction were 40% too large, as compared with computations derived from a validated and process-based model. With the inclusion of the advection correction term, we observe convergence between measured and calculated values of N-e on hourly, daily and yearly time scales. We cannot, however, conclude that inclusion of a one-dimensional, vertical advection term into the continuity equation is sufficient for evaluating CO2 exchange over tall forests in complex terrain. There is an indication that the neglected term, u}(partial derivative cmacr/partial derivative x), is non-zero and that CO2 may be leaking from the sides of the control volume during the winter. In this circumstance, forest floor CO2 efflux densities exceed effluxes measured above the canopy.
