基于泰勒级数迭代的重力归一化总梯度

针对重力归一化总梯度常规计算方法中向下延拓计算的不稳定性，提出了基于泰勒级数迭代法进行快速稳定向下延拓的重力归一化总梯度场( GH 场) 计算模式。水平圆柱体模型试验表明: 该方法的GH 场表现为单一高值特征，而且极大值位置与模型体质心位置相吻合。非均匀( 含油气) 球冠模型表明: 在油气含量较低时，Fourier 级数法和Fourier 变换法无法识别出“两高夹一低”的储油特征，而本文方法可以清晰地识别。这充分说明基于泰勒级数迭代的重力归一化总梯度计算方法不仅提高了计算结果的稳定性，而且还提高了异常的分辨率。

Calculating normalized full gradient of gravity anomalies using Taylor series iteration

In order to improve the instability problem of downword continuation in its conventional calculation， we propose a method of calculating normalized full gradient of gravity anomalies based on Taylor series iteration method for rapid and stable downward continuation． The experiment of infinite horizontal cylinder proves that the GH field characteristics of this method perform a single high value and the maximum position coincides with the location of model centroid． The asymmetric-density oil storage sphere-coronal model prove that this method has still distinguished the characteristics of“one less value between two high values”． The method of Fourier series and Fourier transform do not when the thickness of the oil storage part is low． These fully illustrate that normalized full gradient of gravity anomaly computation method based on Taylor series iterative not only enhances the stability of calculation result，also improves the anomaly resolution．