【作者】 贺哲；

【导师】 沈桐立； 马镜娴；

【作者基本信息】 南京气象学院 ， 气象学， 2004， 硕士

【摘要】 本文详细介绍了以变分法为基础，以伴随码技术为核心的MM5伴随模式，并将常规资料和非常规资料应用于该模式进行了同化问题的研究。该研究是用一个个例进行对比分析，我们选用1998年7月29日00时至30日00时发生于长江流域的一次暴雨过程。对比研究结果表明，MM5伴随模式能有效同化常规观测资料，调整初始场。使用调整后的初始场对降水及其它物理量场的预报均有所改善。其次进行了将MM5伴随模式用于修正模式地形参数的试验研究。利用MM5伴随模式对常规资料和非常规资料云导风进行同化，通过不同的试验研究，结果发现，在伴随模式中加入非常规资料云导风能够比仅使用常规资料更有效地修正模式地形参数，将此地形场作为模式地形所做预报结果比只使用常规资料能够有效改善。更多还原

【Abstract】 In this paper, the MM5 adjoint-model that is based on variational method with adjoint codes technique as the core is introduced, on the basis of which, conventional observation data and unconventional observation data is applied into this Assimilation System for the purpose of assimilation research. First, choosing the conventional data of the heavy rain case that occurred in the Yangtze river drainage basin on July 29th, 1998 and putting it into the Assimilation System to study the assimilation effect. The result shows that the MM5 Adjoint-model Assimilation System can effectively assimilate the conventional data and adjust the initial field. The forecasting of meteorological element fields and precipitation is improved by use of the adjusted initial field. Second, the experiments of modifying the terrain parameter by means of the MM5 Adjoint-model Assimilation System are carried out. In these experiments, conventional observation data and unconventional data (cloud-derived wind) is used. By comparison of the results of different experiments, it is known that the using of unconventional data in the adjoint assimilation model will correct the topography more effectively than using only the conventional observation data. The forecasting conducted on the basis of this terrain field can be improved by contrast to the results using only conventional data.更多还原