The eddy covariance method is now widely used for estimating the net ecosystem CO2 flux (NEE) in various ecosystems. However, the spatial representativeness and scaling-up of such measurements are of the most critical issues. We investigated the effects of footprint on the CO2 flux measurements by the tower-sensor based on airborne/satellite remote sensing imagery. The spatial distribution of CO2 sink/source was derived from the spectral imagery and semi-empirical models. The flux value at the tower-sensor was estimated from the sink/source map combined with a footprint model. Comparison between estimated and measured flux values suggested that footprint effects cannot be negligible when the sink/source distribution is not uniform.