Abstract:Almost all numerical prediction models have systematic errors. Although the statistical correction method has achieved some success in reducing the wind speed deviation at individual stations, it is still urgent to develop a high-resolution wind forecast bias correction method under complex terrain based on grid points. In this study, a fusion prediction and correction technique for high-resolution wind field at different altitudes in Winter Olympic Games area under complex terrain is proposed. Firstly, the statistical bias correction method is used to obtain the 1-12 h average systematic deviation of wind at each station by using the wind observation data of 133 automatic weather stations in the Winter Olympic Mountain Competition area and the high-resolution wind forecast data of RMAPS-RISE system. Then the wind of CMA-BJ model is modified by downscaling and optimized by the grid deviation correction coefficient. Finally, the optimized high-resolution wind is modified by incorporating the latest surface observational data. The results show that this method greatly reduces the systematic deviation of wind speed and the average absolute error of wind speed forecast. The reduction rate of the average absolute error and root mean square error of wind speed within 12 hours is more than 40%. With appropriate modification, this method can also be applied to the bias correction of other variables.