Evaluation of various downhole data reduction methods for obtaining reliable V-S profiles

Abstract

The downhole method has been widely used to measure in situ shear wave velocity profiles for the seismic response analysis of geotechnical sites. To analyze the downhole data, the direct and interval methods are mostly used in practice. In this study, the modified interval method based on a straight ray path and the inversion method based on Snell's Law ray path were introduced to improve the quality of the wave velocity profiles evaluated by the downhole seismic method. Various synthesized wave velocity profile models were developed to perform the parametric study and the arrival times were determined by a forward modeling scheme based on Snell's Law. By comparing the velocity profiles obtained by four different data reduction methods with actual velocity profiles, the accuracy and limitation of various data reduction methods were assessed. The direct method was difficult for evaluating the detailed velocity profiles, and the interval method was found to provide severe errors, particularly when a stiff layer is located beneath the soft layer. The modified interval method provides reliable results, except when a strong soft-to-stiff contrast exists. Snell's Law ray path method provides the most reliable velocity profiles. Finally, in situ downhole seismic tests were performed at three sites, and the importance of considering the ray path in the data reduction was emphasized by comparing the reduced shear wave velocity profiles with crosshole and standard penetration test results.