Skip to Main Content
In this paper, in order to evaluate the simulation capabilities that between the four boundary layer (BL) parameterization schemes, we employ WRF model to simulate the low-level wind of a wind farm which with a complex terrain in Xinjiang for February 2008. The results show that, four BL schemes forecast predominant wind direction and the trend of wind speed accurately but these simulation results have the upper systematic bias. It is found that the accuracy of simulation stability is an important factor in wind speed prediction by studying the distribution features of wind speed simulation errors and simulation results of atmospheric stability. The spectral analysis of wind speed time series indicates that these BL schemes can simulate the wind speed oscillations accurately. In general, there are no pronounced differences between the four different BL schemes simulation results, however, the simulation results from the QNSE are the closest to observation. We conclude that the WRF outputs can be considered for wind power prediction in Xinjiang with arid climates, scarce vegetation and complex terrain of China.