Abstract:Winter sports are closely related to meteorological conditions, which is one of the most important factors for the success of Winter Olympic Games. According to the meteorological support requirements of Beijing Winter Olympic Games, CMA Earth System Modeling and Prediction Centre has preliminarily established a high-resolution regional ensemble prediction test system based on the multi-scale blending (MSB) initial condition perturbation method and lateral boundary condition (LBC) perturbation method, and carried out continuous tests for the same period corresponding to the 2022 Beijing Winter Olympic Games. The statistical results of the preliminary test show that the root mean square error of the ensemble mean values of isobaric surface elements at high, medium and low layers of the main forecast variables is basically less than or equal to the control forecast error, which reflects the advantages of the ensemble mean over the single deterministic forecast. The forecast effect of surface element wind and precipitation is good, but the deviation of 24 h temperature forecast is 2 ℃ higher, which is far from the accurate Winter Olympic Games meteorological support. The high-resolution regional ensemble forecast synoptic analysis of two cold waves and strong wind processes during the test period shows that the ensemble forecast products can accurately describe the main distribution characteristics of ground temperature, cold wave movement process and precipitation forecast, and provide valuable probabilistic forecast information for forecasters, such as cold wave standard 24 h varying temperature forecast and gale forecast. Based on the diagnostic method, the weather element ensemble forecast products that have an important impact on the operation of Winter Olympic events and athletes’ performance such as visibility, strong wind and precipitation phase are developed. The preliminary test results show that the choice of different ensemble members is sensitive to the visibility forecast, and has a certain forecast ability, but the forecast range is too large, and the value is low, which needs to be further improved. The distribution of gust forecast is consistent with the actual large-value area. The precipitation phase forecast is consistent with the observed distribution, especially the boundary of rain and snow, and the location ranges of rainfall, sleet, snow and ice particles are reasonable, which further improves the meteorological support ability to the Beijing Winter Olympic Games.