Abstract: Based on the continuous observations of land-atmosphere flux exchanges, meteorological and biological elements in 2008 from the Jinzhou agricultural ecosystem research station, and by introducing a dynamic surface albedo (α) parameterization scheme that takes the leaf area index (LAI), the solar altitude (hθ) and the surface soil water contents (SWC) into account, the BATS1e model was modified and used to investigate effects of dynamic surface albedo parameterization on flux exchanges between the maize farmland ecosystem and the atmosphere. The results showed that α was able to be dynamically simulated and its simulation error and the Nash-Sutcliffe efficiency factor (NS) decreased obviously and increased by 0.65, respectively, which increased the simulation accuracy of the thermal action in land-atmosphere flux exchanges such as the net absorbed solar energy flux which was improved most obviously among the radiation elements and whose annual improved magnitude was 81772 kJ/m2 making up about 1.7 percent of the annual global solar radiation and the surface soil temperature whose annual and monthly mean improved magnitudes were 0.62 K and above 1 K in most of months, respectively. In addition, the improved model realized the dynamic change of main canopy structure parameters reflecting surface characteristics such as LAI and the vegetation coverage and made flux exchange processes more close to the real for example sensible heat flux and latent heat flux whose simulation accuracies were also improved with their NSs increasing 0.516 and 0.1 as well as the explanatory ability of simulations to observations increased by 6% and 9% in the growing season, respectively, which were larger than in the no growing season.