克拉玛依大风数值预报中的重力波拖曳方案应用研究

新疆克拉玛依地区位于背风坡,长期受翻山气流形成的局地大风影响。该地区日常大风预报业务主要依托克拉玛依气象局的精细化数值预报系统,为进一步完善该系统的预报效果,本文开展了重力波拖曳参数化方案应用试验,对比分析了三组试验方案:不开启重力波拖曳、只开启外层区域重力波拖曳以及两重嵌套设置下均开启重力波拖曳的结果。试验结果表明:在克拉玛依大风预报个例中,对于山区地形背风坡处的克拉玛依站,不启动重力波拖曳会导致风速偏大和起风时间较早;开启外层重力波拖曳方案可以延迟克拉玛依站起风时间,两重区域均开启可以进一步延迟该站的起风时间,使起风时间更接近实况;对于非山区地形背风坡处的站点,在内层开启重力波拖曳会导致一定的负效果;对内层区域的统计检验结果表明不开启重力波拖曳,会使地形复杂区域风场预报存在一定的正偏差,在外层开启重力波拖曳会使正偏差有一定改善,而两重嵌套均开启重力波拖曳方案会导致非地形陡峭区域的负偏差增加。综合全区域内大风预报检验结果表明,只在外层开启重力波拖曳会获得最小的误差和较好的预报效果。

Application of gravity wave drag scheme in numerical weather prediction of strong wind in Kelamayi

Kelamayi has been affected by strong wind for many years,due to its leeward slope location.At present,the Kelamayi Meteorological Bureau has developed a high-resolution numerical prediction system,which can provide reference for strong wind forecast.In order to further improve the prediction effect of the system,a gravity wave drag (GWD) scheme application test has been carried out,schemes are designed.This includes shutting down the GWD in both domains of the model,opening the GWD in the outer domain while shutting down the GWD in the nested domain,and opening the GWD in both the outer and nested domains.The results show that,for the strong wind forecast of Kelamayi,located on the leeward slope,shutting down the GWD in both domains will trigger the downslope winds much earlier than the observation,while opening the GWD in the outer domain will delay the downslope winds to some extent,and opening the GWD in both domains will further delay the downslope winds and minimize the wind speed,so that the results are closer to reality.Although the GWD scheme can effectively improve the time and magnitude of the local gale forecast in Kelamayi City,for the station located on the non-mountain terrain,opening the GWD in the nested domain will yield a certain negative effect.In addition,the statistical results show that opening the GWD in the outer domain can significantly decrease the positive bias of wind forecast and decrease the RMSE,while opening the GWD in both the outside and nested domains will add a negative bias.According to the general verification result,the minimum error and a superior prediction effect can be obtained by turning on the GWD in the outer domain.