变分同化方法在Lorenz系统中的简单应用研究

Simple Application of Variational Four-dimensional Assimilation in Lorenz System

利用Lorenz模式作变分同化数值试验 ,通过对一个简单系统的讨论 ,介绍四维变分同化方法。对初值敏感性和观测点的个数及观测值作了对比试验 ,发现随着模式对初值敏感性的增加 ,同化效果会越来越差 ;观测点越少 ,观测值误差越大 ,这些都会影响同化效果 ,甚至导致同化失败。

The adjoint technique is used in the variational data assimilation using the famous Lorenz model. The numerical experiment shows that the four dimensional assimilation has some relation with the model's predictability, with the increase of error of initial guess field, the effect of the assimilation model become even worse, until the assimilation process fails completely. In the time window of assimilation, if the points of observation are enough, the errors of the observation will not affect the effect of assimilation very much, this will be beneficial for the satellite and radar data assimilation used in operational weather prediction model.