【作者】 郑祚芳；

【导师】 沈桐立；

【作者基本信息】 南京气象学院 ， 气象学， 2002， 硕士

【摘要】 资料同化在数值预报中的重要性正受到广泛的关注，鉴于目前常规观测网在时空分辨率上的局限性，采用分辨率较高的卫星探测资料成为一条重要途径。本文以1996年6月28日～7月4日发生在江淮流域的两场暴雨过程为例，依据统计特征分别以常规分析场、正涡度大值区、高低空急流为标准设计了三套方案对卫星反演湿度场进行质量控制，引入多重网格法来加速同化求解过程。利用MM5模式进行的数值试验表明：(1)经过质量控制后，预报场得到较大的改善，特别是前6小时降水预报准确率有明显的提高；(2)质量控制方案的优劣对预报效果有较大影响，选择合理的控制标准是重要的；(3)多重网格方法具有收敛速度快、精度高的优点，用于同化过程能提高模式预报时效；(4)自适应网格技术的引入有助于提高数值模式的局部分辨率，改善预报效果。更多还原

【Abstract】 Nowadays, greater attention has been drawn to the importance of data assimilation in numerical prediction. It is necessary to use the satellite image data that has a high resolution due to the time-space limit of the conventional sounding system. In this paper, the numerical simulation by the MM5 model about twice heave rainfall during 28 June -4 July 1996 in Yangtze-Huaihe river basin are studied. By using the multi-nested grid method to speedup the simulation process, three schemes are developed to control the quality of satellite retrieved humidity, based on conventional analysis data, the field of positive vorticity, high/low level jet, respectively. The results may be concluded as follows: (1) The simulation accuracy is improved greatly under quality controlling, especially about the 6h precipitation; (2) The quality control is critical to the forecasting effect, it is important to select the rational scheme of quality control; (3) Using the multi-nested grid method can improve the period of validity because it has higher speed and accuracy than SOR method; (4) The application of adaptive mesh model is helpful to improve the resolution of numerical model in local.更多还原