The Interpretation of Moment Tensor Inversions

Abstract

The moment tensor formalism provides a concise means of expressing the body force equivalents for a variety of simple point sources; consequently, it is ideally suited for use in the forward problem of generating synthetic seismograms for most sorts of natural and artificial sources. Furthermore, this straightforward approach for the computation of synthetic seismograms would indicate a relatively simple inverse problem. However, it is the very generality of the method which poses potential problems when trying to interpret the relationship between the elements of the moment tensor determined from the inverse problem and the physical reality or model of the source. The very same a priori knowledge and assumptions that permitted a simple construction of the moment tensor will provide the stumbling block that precludes a facile implementation of an automated analysis system. The root of the problem lies in the inability of a rote technique to duplicate the complex chain of reasoning which led to the imposition of the a priori constraints and the adoption of an idealized earth model. In those cases where the inversion has been limited to determining the magnitude and orientation of the best double couple, the difficulties alluded to above are not restrictive. In the more complex situation wherein one seeks to compare source models or to investigate the existence of a particular source type, these constraints may present quite formidable obstacles.