Impact Of Background Fields And Boundary Layer Schemes On Surface Layer Wind Prediction

The prediction of surface layer wind is an important concern for the utilization of wind energy resources.The initial-and boundary-value of the regional model come from background field,and the description of surface layer turbulence process requires a boundary layer(BL)scheme.Therefore,both background filed and BL schemes have an important impact on the prediction of surface layer wind.To evaluate the effect of BL scheme and background fields on wind prediction in western Inner Mongolia,WRF model combined with the GSI assimilation system was used to conduct two sets experiments with different background fields and different BL parameterization configurations,and observations of 70 national stations and 18 wind towers in the concerned region were used to assess the results of the hindcasts.GFS and T639 data were compared on wind field forecasting,and the results reveal that: 1)The prediction of the two background data reproduced the wind evolution in western Inner Mongolia area during the simulation period.The predicted 10 m wind speed is higher than the observations,and the prediction of wind direction has a small positive deviation.2)Through the gap between the two background fields is very small,the prediction of GFS on wind at 10 m and 70 m is slightly better than T639.3)The hindcasts are significantly affected by resolution and forecasting time of the model,.the forecasting performance of 4 km is better and 12 km,The performance of forecasting time is(12~35)h >(36~59)h >(60~83)h.4)The forecasting of model will be affected by the distribution of complex terrain,because the stations with large errors mainly locate in the low-lying areas of the southern Yinshan Mountains.Experiments on seven different BL schemes reveal that: 1)The prediction of WRF model can reproduce the evolution and phase changing of the wind speed and wind direction.There are clear differences between 10 m wind speed predictions,YSU get the significantly better prediction of wind speed at 10 m,the performance of ACM2 is better by compared with other BL schemes on wind speed prediction at 70 m,the simulated wind speed is larger than the observations,the model is poorer for low wind speed prediction.The wind direction at 10 m and 70 m simulated by all sets of hindcasts has a slightly clockwise deviation to the observations.2)The performance of all the BL schemes for wind segment is(3~15)m/s >(15~25)m/s >(0 ~3)m/s.3)The diurnal variation of the wind speed and direction can be simulated in all BL schemes: wind speed increases with the time and wind direction rotate clockwise in the daytime;wind speed decreases and wind direction rotate counterclockwise in the dusk and night.4)TEMF,YSU and ACM2 performance better than other BL schemes on wind shear index prediction,the atmospheric stability has a significant effect on the forecasting of lower level wind.In summary,better prediction of wind can be obtained by both GFS and T639.The forecast results are ideal when the model resolution is 4 km and forecasting time is 12~35 hours.YSU get the significantly better prediction of wind speed at 10 m,though the differences of the BL schemes on the prediction of the 70 m wind field is not obvious,and the performance of ACM2 is better by compared with other BL schemes.TEMF,YSU and ACM2 performance better than other BL schemes on wind shear index prediction,the atmospheric stability has a significant effect on the forecasting of lower level wind.