Observation Investigation On Fine Features Of The Near Surface Wind In Winter Over Beijing Area And Evaluation On Performance Of GRAPES-Meso Model

Under the background that the 22 nd Winter Olympic Games will be held in Beijing in 2022 and the previous host countries adopt the technical route of combining the observation law of specific topography with the development of high resolution mesoscale numerical model,this paper uses the near-surface observational wind data of 208 stations in Beijing area from 2009 to 2018 to observe the fine characteristics of Beijing,four typical region and Yanqing Alpine stations.Based on the observational characteristics,the wind forecast characteristics of GRAPES-Meso 3km model from 2016 to 2018 are compared with the observational results,and the forecast deviation of the model is evaluated.Finally,the possible causes of the model forecast deviation are analyzed.The specific conclusions are as follows:(1)The average near-surface wind speed in winter in Beijing is higher in the West and Northwest Mountainous areas than in the plains and weakest in the city.The prevailing wind in winter is mainly northwest and northerly.It reflects the basic distribution characteristics and winter characteristics of near-surface wind in Beijing under the influence of topography and urban underlying surface roughness and cold air in Northwest China.The diurnal variation of winter wind speed in four typical regions(mountain area,transition area,plain and urban area)shows that the wind speed is higher in the day than night and got single-peak wind speed in the afternoon.The average daily wind speed of two higher mountain stationsin Yanqing Alpine area is several times higher than that of two lower mountain stations,and the diurnal variation law of wind speed of the two high stations is completely opposite to that of the other two stations.This feature of gradient wind speed in the competition area is related to the diurnal variation of turbulent activity in the boundary layer and the downward transfer of free atmospheric momentum in the lower layer to the boundary layer.(2)The GRAPES-Meso 3km model can generally reflect the climatic characteristics of the observed mean winter wind speed in Beijing and the dominant wind direction characteristics,but the wind speed predicted by the model is larger overall,and the prediction of the model in mountainous areas is more unstable than that in plain.The model performs well in terms of the variation characteristics of average wind speed in four typical regions,at the same time,there are systematic positive deviations in the wind speed forecast,and the deviation in night time is larger,the peak time of wind speed forecast is 2-3 hours ahead.The diurnal variation characteristics of winter wind speed and dominant wind direction at four alpine stations in Yanqing Alpine area are not well forecasted.(3)The forecast deviation of wind factors for Beijing region(mainly transition area and mountainous area)is related to the deviation of model terrain height and observed terrain height,while the stability of wind speed prediction is mainly related to other parameters of the model itself.For the wind factor prediction deviation of the four high-altitude stations in Yanqing Alpine area,besides the great difference in the model topographic height between observation in the four stations,the model still shows the characteristics of the boundary layer in the two high-altitude stations,while the observed topographic height of the two stations is correlated outside the boundary layer.