摘要
由刘元龙和王谦身提出的用于反演重力资料以估算地壳构造的压缩质面法，是用二维方柱体来近似地质体，而为了得到反演问题的解，又将二维方柱体近似地表示为压缩质面，使矩阵分析方法得到实际应用。利用这种压缩质面法，可根据实测重力异常资料反演莫霍界面的分布，而在已知几层密度差异的条件下，还可反演多层密度界面，(如莫霍界面和康腊界面）的分布。
The kernel content of the method offered in this paper to inverse gravity data is still the compressed mass-surface method, in which two-dimensional rectangular prisms approaching the geologic bodies are approximately represented by compressed mass-surfaces, so that the matrix analysis method can be used to practise. However, being improved on many respects of original method, present method has fitted into inversion of gravity data in unequal station intervals, speeded up convergence procedure and enormously raised the computation accuracy.In order to obtain the accurate distribution of density boundary, three steps have been principally provided to inverse gravity data: The first is to use a formula of infinite horizontal slab as an initial inversion, called a pre-adjustment of boundary depths, the purpose of which appears only to put the compressed mass-surfaces on more reasonable positions for subsequent inversion through the compressed mass-surface method; the second -to inverse gravity data by the compressed mass-surf ace method of itself, which can obtain the approximate result of boundarygthe third is to obtain the residual difference of gravity anomalies through direct gravity computation and to modify boundary distribution by using the compressed mass-surface method, and it is called adjustment of boundary that can be repeated many times until reaching the requirement of accuracy. Since the step of the pre-adjustment has been added into computation and the compressed mass-surface have been put on the most reasonable positions according to different cases, more ideal results have been given by the offered method.In this paper, the results of trial computation are illustrated as a comparison with the original method, and some problems of the method are discussed.
