航空重力梯度数据向下延拓的正则化方法及参数选取研究

围绕航空重力梯度测量数据处理的向下延拓问题，利用不同的正则化算法和参数选取方法对向下延拓的数学模型分别求解。通过梯度数值的模拟，联合EGM2008位系数计算的大地水准面上的重力场参数，对向下延拓结果进行检核。模拟数值计算结果表明，截断奇异值正则化（TSVD）方法结合广义交叉验证（GCV）参数选取方法能够实现稳定的重力场信号向下延拓，精度和稳定性都优于其他几种方法的组合。

A Study of Regularization Method and Parameter Selection of DownwardContinuation of Airborne Gravity Gradient Data

This paper compares and analzes different regularization algorithms and parameter selections methods applied to the downward continuation of airborne gravity gradient data. Validation is achieved by comparing the results of the numerical simulation with those directly calculated by EGM2008. The numerical results show that a combination of the truncated singular value decomposition (TSVD) regularization method and the generalized cross validation (GCV) parameters selection method is most robust, and that this combination method is preferable to the other methods with respect to accuracy and stability.