使用不同的天气尺度初值作强对流暴雨预报的差别及其原因

分别用暴雨开始前和暴雨刚开始两个探空时次的天气尺度初值场作1982年6月19～20日鄂东特大暴雨过程的中尺度模拟预报.前者相当成功,后者完全失败.探究其原因说明中尺度数值预报可以用天气尺度初始场的一个重要条件是模式初始时刻在关注的区域及其附近不存在即将造成或已经造成强降水的强中尺度扰动.如若不然,则可能由于模式起转(Spinup)的延误,实际存在的中尺度扰动对天气尺度变化的强大反馈作用不能体现,会使天气尺度的预报从一开始就显著歧离实际,其带来的结果不是中尺度时、空细节的预报误差,而是全区域性的预报失败.这个试验还进一步证实伴随暴雨的强非绝热反馈作用是暖切变型梅雨锋发展和准静止持续乃至周边区域形势维持的最重要的直接因子.

The Difference and Reason of the Numerical Predictions of Severe Convective Heavy Rain Using Synoptic Scale Initial Data in the Different Practical Initial Situations

An extremely heavy rain event in east Hubei Province during June 19-20,1982 is simulated by using respectively the synoptic scale initial data of two sounding times in mesoscale model. The forecast is quite well with the initial data time early before the heavy rain and fail with that after the beginning of the heavy rain. The reason of this surprising result is probed. And then it is recognized that if we hope to succeed in using synoptic scale initial data in mesoscale numerical prediction, as Anthes had optimistically argued in 1984, an important prerequisite should be satisfied. This is that there is no strong mesoscale disturbance which will soon or already initiate up intense precipitation in the concerned region and its surrounding at the model initial time. Otherwise, the significant feedback effects of the practically existing mesoscale disturbance to the synoptic scale evolutions cannot be realized due to the delay of model spinup. Probably the synoptic scale prediction will thus go astray far away from the fact. The consequence will be not only the error of mesoscale details temporarily and spatially, but also the failure of the wholly regional forecast. The experiment also confirms that the feedback of strong cumulus heating is the most important direct factor in developing and stationarily maintaining the worm shear-line type Mei-yu front and even maintaining the environmental circulation situation.