中文摘要：

以如何提高黄海海雾数值模拟初始场质量为研究目的，利用WRF模式及其先进的3DVAR同化模块，设计并构建了循环3DVAR同化方案。以2006年3月6～8日的1次大范围黄海海雾过程为研究对象，利用该同化方案进行了一系列WRF数值模拟对比试验。模拟结果显示，循环3DVAR同化方案能有效改进黄海海雾数值模拟初始场质量，主要体现在增加低层大气温度层结构的稳定性与改变大气边界层下层的风场结构，从而导致海雾的模拟结果显著改善。研究结果表明进行海雾数值模拟时，必须高度重视其初始场质量。

英文摘要：

From the view of improving initial conditions of sea fog modeling over the Yellow Sea, a cycling 3DVAR data assimilation scheme is designed based on WRF model and its 3DVAR module. By using this scheme, a series of WRF numerical experiments of a heavy sea fog event over the Yellow Sea from 6 -8 March, 2006 are conducted to evaluate its performance. The results show that the scheme can greatly improve the numerical modeling initial conditions, which results in a much better simulated sea fog coverage compared to satellite observations. The improvements of initial conditions are mainly embodied in strengthening layer stability of the whole marine atmospheric boundary layer （MABL） and modifying winds of lower MABL. This study indicates obviously that it is very important and necessary to improve initial conditions, as much as possible, for sea fog numerical modeling over the Yellow Sea.