Common assumptions made in the interpretation of seismic refraction data are that the refractor has a uniform seismic velocity and that there is negligible dip on the refractor boundary. Recent developments make it clear that assumptions such as these have little relevance to the realities of crust and upper mantle structure. The consequences of making these assumptions, for example, in a time-term analysis, are, therefore, potentially serious.
In this paper, data simulation followed by time-term analysis is applied to a variety of complex crustal models to examine the implicit constraints of the time-term method. Although disconcerting in some respects, the results are of practical value in the interpretation of seismic refraction data and have been applied to some data acquired in a recent British explosion experiment. It is concluded that directionally dependent variations in crust and upper mantle structure may be of considerable importance; the results of all crustal studies, past and future, should be interpreted with this in mind.