背景场资料与参数化方案对暴雨预报的影响

The Impacts of Background Fields and Parameterization Schemes on Torrential Rain Forecast in GRAPES-meso

  • 摘要: 利用GRAPES-m eso模式和T213资料,对2007年7月18日发生在我国四川盆地和华东地区一次大暴雨过程进行多组数值试验,以分析侧边界资料、驱动资料的垂直分辨率、模式积分区域、云物理参数及边界参数对GRAPES-m eso模式降水预报影响。试验结果表明:(1)侧边界资料对模式降水预报结果影响较小,驱动GRAPES-m eso的全球模式产品质量提高,降水预报结果越好;(2)驱动资料垂直分辨率的高低对降水预报结果影响较大,分辨率越高,预报能力越强,反之越弱;(3)模式积分区域对降水预报结果也有明显影响,区域越大,降水预报未必总是最好;(4)物理过程和边界参数试验表明,W SM6方案与KF eta方案组合的24小时降水预报与实况更接近。

     

    Abstract: Based on GRAPES-meso model and T213 data, a series of numerical experiments are carried out to trail a rainstorm process in Sichuan and East China during 18-20 July 2007 in order to analyze the impacts of lateral boundary condition, vertical resolution of T213 data, the model domain of integration, physical process and boundary parameter on the rainfall forecast by GRAPES-meso model.Through four groups of comparative experiments, the main conclusions are as follows:(1) There are only slight impacts of lateral boundary information on this torrential rain forecast.With the qualities of the background fields used in GRAPES-meso model higher, the forecasting accuracy of this rainstorm process are better. (2) Vertical resolutions of T213 data in GRAPES-meso model are very important to the accuracy of this weather process.The higher the vertical resolution is, the better the quality of rainfall forecasting is.And, contrariwise, the weaker the quality is. (3) When the integration domain in GRAPES-meso model is different, the results of precipitation simulations are obvious different.The bigger the model domain of integration is, not always the better the results of simulative rainfall are. (4) The test results using different physical process and boundary parameter show that, the forecast of 24-hour accumulated precipitation is close to the observed precipitation when using WSM 6 scheme and KF eta scheme.

     

/

返回文章
返回