【摘要】 论述了数据同化中的伴随方法与泛函分析中的伴随算子的关系 ,指出经典的拉格朗日乘子法可以作为伴随方法的理论基础。关于伴随方法应用中只能利用模式的伴随的观点 ,作者认为是有疑问的 ,并提出了新的观点。所作的数值模拟实验表明 ,方程的伴随有时明显优于模式的伴随 ,因而说明方程的伴随也应该引起足够的重视。更多还原

【Abstract】 The relations of the ad joint method in data assimilation and the ad joint operator in functional analySis werestudied. It is POinted out that the classical Lagrange multiplier method is one of the theoretical fundament of the adjoint method in data assimilation. The author holds it is not reasonable that one can only use the ad joint of model indata assimilation. The simulated numerical experiment shows that the ad joint of equation is sometimes superior tothe adjoinic f model.更多还原