江淮锋面和华南暖区两次暴雨过程可预报性对比

Comparison of the model predictability between a front rainstorm in Yangtze-Huaihe river basin and a warm-area rainstrom in South China

以2008年6月9—10日江淮地区的锋面暴雨和2008年6月6—7日华南地区的暖区暴雨为例,采用模式试验的方法,研究了这2个不同地域不同类型的暴雨的模式可预报性的差异.控制试验的结果表明,2个地区的暴雨都可以用WRF模式得到较好的模拟再现.通过在控制试验的初始场上对温度场和风场添加高斯随机扰动误差构造敏感性集合成员,结果表明初始场的微小误差在24h内使得华南暴雨与江淮暴雨的模拟结果都发生较大改变,但华南暴雨的误差增长快于江淮暴雨,导致华南暴雨模拟结果发生更大的改变.通过对集合离散度的分析表明,华南暴雨与江淮暴雨的离散度都随积分时间延长而不断增大,但华南暴雨的集合离散度增长更快,华南暴雨的集合离散度在模式各层上都远大于江淮暴雨.从误差增长和集合预报的角度讲,华南暴雨的模式可预报性比江淮暴雨的模式可预报性差.

The difference of the model predictability between front rainstorm in the Yangtze-Huaihe River Basin and warm-area rainstorm in South China is investigated through numerical model sensitivity experiments. The control experiments indicate that the rainstorms in the Yangtze-Huaihe River Basin and South China can be reproduced by WRF model. The experiments were designed in which perturbations consist of random errors with Gaussian noise were added to the initial conditions, including temperature and wind fields. The result shows that small errors in initial conditions make the simulation results change a lot after 24 h both for front rainstorm in the Yangtze-Huaihe River Basin and warm-area rainstorm in South China, but the forecasting error grows faster in south china than in the Yangtze-Huaihe River Basin,which lead to the bigger change in simulation results. Furthermore, we found that the dispersion grows with time, and the magnitude of dispersion in south China grows faster than that in the Yangtze- Huaihe River Basin. In addition,the model predictability of warm-area rainstorm in South China is less than that in the front rainstorm in the Yangtze-Huaihe River Basin.