Comment on How Necessary are Large-Scale Refraction Experiments?

Abstract

Fundamental requirements for refraction surveysIn commenting on the foregoing communication of Francis, it seems necessary to start with a fundamental statement of the requirements and possibilities of refraction surveys, and then to consider the practical techniques as special cases. The general requirements are as follows:(a) A clear boundary must exist such that materials below the boundary propagate seismic waves within a higher range of velocity than those above.(b) Waves from several sources, refracted through the materials below the boundary, should be recognized at several detectors.Given sufficiently dense coverage, it is possible to determine the two-dimensional distribution of propagation velocity for the materials immediately below the boundary, and all of the ' delay times ' which describe the transmission of the waves through the upper layer. Types of refraction surveyThe complete solution would require very dense coverage and, if all connections were to be used, an impracticably complex technique of interpretation. The following two techniques for reducing the total complexity are therefore used:(a) Linear profiles. In a reversed linear profile, wave fronts passing under the central section have most of their propagation paths in common with each other, and this enables the propagation velocities below the layer, and the delay times for transmission to the surface, to be determined separately (Hagedoorn 1959). When an abrupt change of structure occurs within the profile, delay times for a single station, looking along the profile in opposite directions, may be different. O'Brien (1968) allows for this difference in his interpretation of the Lake Superior experiment, and uses it to estimate the velocity in the upper layer.(b) The time-term approach. In this case the distribution of shots and detectors may be two-dimensional. The solution is derived on the assumption that the material below the boundary has a uniform propagation velocity, and that the delay time for any given survey point is the same for all the ray paths which begin and end at that point. This requires careful definition of the quantities involved.We start with the general refraction equation t i j = U , +~~+ A~~/ U~ 227

An Interpretation of First-Arrival Data from the Continental Margin Refraction Experiment

“…Whatever one's opinion of arguments such as those of Willmore (1969) and Francis (1969) regarding the relative merits of the refraction and reflection methods for crustal studies, it is evident that crustal structure is far too complex to be accurately defined by simple refraction experiments requiring for solution assumptions such as uniform plane layering; indeed it cannot be fully described by even the most sophisticated interpretations of refraction data alone. To this extent the refraction and reflection techniques may be regarded as complementary and CMRE was designed as far as possible to make use of both.…”

An Example of The Unified Method of Interpretation for Crustal Seismic Data

“…Even though most of the quantitative information about the interior of the Earth has been gleaned from painstaking studies of seismic wave propagation, a recent issue of this Journal has carried a discussion in which the value of seismic refraction studies of the crust was questioned (Francis 1969) and, in the opinion of the present authors, the reply to a number of these points (Willmore 1969) is in need of some amplification.…”

“…The authors would also like to emphasize the importance of the information carried by later arrivals, especially those near critical points, in affirmation of the statement by Francis (1969) that Russian workers are properly concerned with the dynamic as well as the kinematic characteristics of wave groups. Willmore (1969) addresses nearly all of his reply to a defence of the time-term approach, and overlooks the possibility that any except first-arrivals will be of much use in refraction interpretations. The reporting of one-layer crustal models, which often results from time-term studies, is certainly-as Francis correctly notes-an oversimplification.…”

An Example of the Iterative Approach to Time-Term Analysis

“…The basis of the iterative method was described by Willmore & Bancroft (1960) in their initial formulation of the time-term approach and later, in condensed form, by Willmore (1969).…”

“…Within such a unified interpretation procedure, a correctly employed time-term analysis could play a considerable part; one notes, for example, that time-term analysis is capable of evaluating lateral variations in refractor velocity, velocity increase with depth, velocity anisotropy and so on. Willmore (1969) has pointed out that many of the criticisms of the time-term method have arisen from a lack of attention to the definitions of the method, and from a failure to adopt the necessary iterative process of solution. The present author would go further than this and say that a majority of the criticisms of the time-term method have arisen as a result of a single experiment, namely the Lake Superior seismic experiment (Steinhart 1964).…”