Traveltime differences in seismic refraction inversion

The application of seismic refraction methods is very widespread in near-surface geological investigations for e.g. hydrogeological, geophysical engineering and geotechnical purposes [1, 2]. In a surface seismic measurement various source types can be used, related to which trigger error may occur during field measurements. If elastic waves are generated by explosion, an approx. 1–4 ms triggering error can occur due to the error of the geophysical blasting cap. Generating the waves with weight dropping or hammer strike, a piezoelectric ceramic or a geophone serves the trigger signal – cycle skipping may occur. These triggering errors can cause problems in the interpretation of measured data and the estimation of the parameters, especially in shallow explorations as the accuracy or the errors of the trigger time have greater importance there compared to that of deeper explorations. This problem was examined earlier in the inversion of three-component seismic VSP surveys in a coal mine using the traveltime differences between the upper and lower geophone-triplet [3]; and in shallow seismic reflection exploration, estimating the shot distortion on each common shot gather and eliminating them by shifting all the traces [4]. In this paper the idea of double-trace data is applied to refraction traveltime data, a concept that was developed by Dobróka et al. [5] for tomographic interpretation. The modified conjugate gradient and SIRT algorithms proved to be effective for solving the problem described above. With this theory the problem of the inaccurate trigger time is attempted to be solved in the field of seismic refraction.