GOES-16大气温度产品资料同化在一次飓风预报中的应用研究

APPLICATION OF GOES-16 ATMOSPHERIC TEMPERATURE PRODUCT DATA ASSIMILATION IN A HURRICANE FORECAST

为评价静止卫星大气温度廓线产品资料同化对飓风预报的影响,以2018年飓风“迈克尔”为例,选用GOES-16温度廓线产品,开展静止卫星资料同化及其对飓风预报影响的研究。首先,通过评估温度廓线产品精度,选取质量较好的高度层并以统计的各层均方根误差作为观测误差用于同化试验;然后,利用WRF-3DVar系统进行不同稀疏化及不同同化频次的循环同化敏感性试验;最后,利用WRF模式开展24 h数值预报。试验结果表明,在飓风“迈克尔”期间温度廓线在200~1000 hPa之间的误差在2 K以内,将水平分辨率稀疏化为模式分辨率的6倍且循环同化频次为6 h时同化该资料对模式的初始场有最为合理的改进,从大尺度环境场上看使模式具备更合理的环流形势,能够有效提高对飓风的路径及强度的预报效果,更准确地模拟降水落区及美国佛罗里达州等降水关键区域的雨强。

In order to evaluate the impact of geostationary satellite atmospheric temperature profile product data assimilation on hurricane forecast,this paper studies the assimilation of GOES-16 temperature profile product data and its impact on hurricane forecast for Hurricane Michael in 2018.First,by evaluating the accuracy of the temperature profile product,we select the higher quality layer and use the statistical root mean square error as the observation error for the assimilation.Then,the sensitivity tests of cyclic assimilation with different sparsity and different time intervals are carried out by using WRF-3Dvar system.Finally,the WRF model is used to carry out 24h numerical forecasting.The experimental results show that the error of the temperature profile between 200hPa and 1000hPa is within 2K for Hurricane Michael.When the horizontal resolution is thinned to 6 times the mode resolution and the cyclic assimilation interval is 6h,the assimilation of the data has the most reasonable improvement in the initial field of the model.From the view of large-scale environmental field,the model has a more reasonable circulation situation,which can effectively improve the forecast of hurricanes’track and intensity.It can also more accurately simulate the precipitation area and precipitation in key areas such as Florida.