In general , 6h accumulated rainfall prediction of mm5 is improved by 3dvar of gps zwd data more significantly than gps pw data although the former improvement is related to the rainfall amount 尽管湿延迟量资料同化对6h累计降水预报的改善在不同的降水等级表现了不同效果，但总体上湿延迟量资料同化对6h累计降水预报的改善程度要高于可降水量资料变分同化。
Consequently , this scheme is feasible and valid . the research enriches the contents of variational assimilation and makes it more extensive application , which increases the quality of initial fields in nwp 因此，将遗传算法应用于四维变分资料同化是切实可行而且有效的，这一工作丰富了变分同化的内容，使变分同化技术作为提高数值预报初始场的质量这一有效技术得到了进一步的应用。
The research on marine atmospheric numerical forecast and data assimilation methods are very important parts in the project of " marine environmental forecasting and natural disasters reducing technology " , which belongs to the national key scientific projects of tenth five - year “十?五”国家重点科技攻关项目“海洋环境预报与减灾技术”中把海洋气象数值预报及资料同化方法的研究列入重点内容。
Through the study of this paper , it indicated that the kalman filter method provides not only the optimal sequential estimation of the system , but also system error information . it is suitable to assimilate the real - time observed data into forecasting model 通过本文的上述研究结果表明： kalman滤波方法不仅提供一个系统最佳状态的顺序估计，而且提供系统误差的信息，适合于将实时观测资料同化到预报模式中。
On the whole , gps pw data assimulation will improve precipitation prediction . 6 , gps zwd assimulation ingests atmospheric information observed by gps into mm5 initial fields more easily than gps pw assimulation so that the initial fields of humidity , temperature , pressure and wind can be modified in a greater degree 6 、 gps湿延迟量资料的三维变分同化比gps可降水量资料同化更容易将gps对大气的监测信息传递给mms模式初始场，使得对模式初始湿度、温度、气压和风场的调整程度大于gps可降水量资料同化对模式初始要素场的调整
The results of numerical experiments , using the four - dimensional variational data assimilation system of the nonhydrostatic mesoscale model mm5 , show that : the tangent linear model and adjoint model conducted by keeping the " on - off " switches the same as the basic state , can provide a good approximation of the first - order information to the nlm perturbation and a good descent direction for the minimization procedure ; switching on and off at every other time step in the kuo cumulus parameterization scheme do n ' t impact the convergence rate of cost function ; the existence of the switches do n ' t impact improvement to the mm5 model rainfall prediction because that not only specific humidity , but also wind , temperature and pressure are assimilated into the model 非静力中尺度数值模式mm5的四维变分资料同化系统进行的数值试验结果表明： “开关”变量保持与基态一致，所构造的切向线性模式能够提供关于非线性模式扰动的一阶近似，伴随模式所计算的梯度值能够为最小化过程提供较好的下降方向；郭氏降水参数化方案中对流每隔一个积分步的交替发生并不影响目标函数最小化的收敛速度； “开关”变量的存在也不影响将风、温度、气压和比湿结合起来同化对mm5降水预报准确性的提高。
In this context a genetic algorithm ( ga ) was applied to the four - dimension ( 4 - d ) variational data assimilation , which provided a kind of new and effective method in the initial field optimization . the theoretical basis and detailed algorithm were introduced in this paper . at the same time , according to the property of the variational problem itself , we designed the rational genetic coding , operators and parameters 本文将遗传算法应用于四维变分资料同化问题，提供了一种新的较为有效的nwp初始场优化方法，给出了相应的理论依据和详细算法，并结合变分问题本身的特点，设计了合理的遗传编码、遗传操作和遗传参数。
Meanwhile the results also showed that the main cause of which the game reanalysis data being better than the ncep reanalysis data is the game reanalysis data including the sonde data of the game but not the different assimilation modes of the two reanalysis data 通过对game外场试验期和非试验期的game与ncep两组再分析资料进行对比分析，我们得到， game再分析资料优于ncep再分析资料的主要原因是game的资料同化系统中加入了包括hubex等气象科学试验的探空加密观测资料。
