The turbulent momentum and sensible heat transfer over land surface have a notable influence on the change of global climate and atmospheric circulation, and Monin-Obukhov similarity function is a most important method to calculate the turbulent momentum and sensible heat flux near the surface, and ascertaining the right bulk transfer coefficient is a most effective way of improving the atmospheric model simulation capabilities. The characteristic of Monin-Obukhov similarity function is analyzed and the empirical formula is fitted, and the changes of bulk transfer coefficients of momentum and sensible heat over grassland with mean wind speed at 10 m high are discussed by using the data of the flux observations over Xilin Gol grassland in Spring 2008. Comparison with the observation values by eddy correlation method shows that the revised Monin-Obukhov similarity function underestimates the momentum flux by 10.8% and over estimates the sensible heat flux by 6.5%, but the typical Businger-Dyer similarity function underestimates the momentum flux by 37.0% and over estimates the sensible heat flux by 16.1%. Under unstable stratification, the bulk transfer coefficients of momentum (CD) and sensible heat (CH) vary with mean wind speed at 10 m high (U) according to the power law, which take the forms CD=0.009U-0.322 and CH=0.184U-1.978 respectively. Under stable stratification, the bulk transfer coefficients are found to increase in the manner of the logarithm law over grassland surface and tend to neutral or nearly neutral values with wind speed increasing. The revised Monin-Obukhov similarity function can significantly improve the accuracy of turbulent momentum and sensible heat flux computed by average gradient data, and the relations between bulk transfer coefficients and wind speed at 10 m high provide the useful parameterization schemes for accurately expressing the transportation characteristics of near surface turbulence.