Abstract: Considering atmospheric, hydrological, biological, and other biophysical processes in a soilplantatmosphere system, upscaling the mechanism model of photosynthetic biochemical processes established by Farquhar from leaf to canopy scale, and improving the canopy layered method, a multilayer, two-leaf model for the estimation of mass transfer and energy exchange between plant canopies and the atmosphere was developed, which combined the vertical variation of nitrogen content in the leaves， and could tell fluxes of water, heat and CO2 above the canopy after calculating the intercepted irradiance, stomatal conductance and photosynthesis of sunlit and shaded leaves separately in each layer. Based on the simulations using this an exponential function of nitrogen level in leaves was established and presented, changing with the height of canopy or the relative cumulative leaf area index and relating to the nitrogen attenuation coefficient of kn. This model was applied to modeling of the CO 2 fluxes for winter wheat canopy at the Yucheng Comprehensive Experimental Station of the Chinese Academy of Sciences during April to May in 2008 with the result that the attenuation coefficients of kn are 0.792 （R2=0.698） and 1.374 （R2=0.728） during the periods of jointing to booting and anthesis to milky maturity, respectively. When using the eddy covariance flux data to verify the validity of the model, it is seen that there is a close relationship between the modeled net ecosystem productivity (NEP) and the measured with R2=0.78, and that the modeled hourly CO2 fluxes indicate better agreement with measurements on clear days than on cloudy days and at nights. The shaded leaves played an important role in the productivity with the contributing rate accounting for 35.7% of the total gross primary productivity (GPP). The leaves of the upper canopy dominated the contribution to the final yield, accounting for more than 80% of the GPP. The modeled total net primary productivity (NPP) was 626.3 g/m2 during the experimental period.