Abstract: Through the different parameterization schemes of the CALMET model, results of simulated wind speed and wind direction of Jiangxi mountain wind field are compared. The results show that the optimal parameters of the fifth set of parameters, namely the CALMET model, are not suitable for detecting kinematic terrain effects. Adjustment of slope flows has smaller effect on wind field above 50 m level. The adjustment of mountain wind field without Froude number and adoption of the O’Brien method can improve the simulation of wind speed at the mountain wind power station. The optimal parameter scheme not only simulates well the trend of high mountain wind power station wind speed, but also produces good results for the simulated distribution of the wind speed segment throughout the year. The CALMET model can effectively simulate the actual wind direction of the actual wind tower, but there is about one azimuth and 7%-8% deviation between the simulated and measured wind direction distribution and dominant wind direction, respectively.