对流云团资料在局地暴雨数值模拟预报中的应用

USE OF CONVECTIVE CLOUD CLUSTER DATA IN THE NUMERICAL PREDICTION MODEL FOR LOCAL HEAVY RAINFALL

使用ＰＳＵ／ＮＣＡＲ中尺度数值模式，对两次暴雨实例进行了数值模拟试验。结果表明：相对于应用地面及高空常规资料作为初始场的控制试验而言，应用实时卫星对流云团资料增强局部湿度场的敏感性试验，可使暴雨及邻近地区对流层中低层气流辐合上升运动增强，并在第４ｈ左右达到最大；早期４～６ｈ内的降水量也较大，比不用这类资料的更接近实际的观测值。如对卫星云图资料增强的湿度场进行初始化处理，则“湿度效应”持续时间大为增长。

Two heavy rain cases are simulated with the PSU/NCAR mesocale model.For control experiments,both the surface and upper air observation are used in the initial fields.For sensitive experiments,the initial fields are the same with the control ones,but the real time GMS satellite convective cloud cluster data are used to modify the initial moisture fields.The results show that (1)because the convective cloud data strengthen local moisture fields,both the convergence and updrafts of the low middle troposphere over the heavy rain center and its peripherial areas intensify and reach maximum in strength about fourth hour in sensitive experiments;(2)because of the use of convective cloud data,the simulated rainfall in the first 4～6 hours is greater than that in the control experiment,and it is closer to the real rainfall;(3)if the modified initial moisture field is initialized with the other initial fields,then the duration of humidity effect is greatly extended.