Full text loading...
-
Non-Uniqueness in Seismic Refraction Analysis
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 24rd EEGS Symposium on the Application of Geophysics to Engineering and Environmental Problems, Apr 2011, cp-247-00019
Abstract
Any seismic refraction analysis is essentially non-unique. in most cases, it is difficult to obtain a true velocity model from traveltimes observed only at the ground surface. Even with a one-dimensional two-layer model, it is impossible to obtain true thickness and velocity without a priori knowledge that the model is two layers. Non-uniqueness is a problem not only for the seismic refraction method but also for most geophysical methods in which underground physical property models are estimated from geophysical data observed at the ground surface. All seismic practitioners and algorithm developers must admit this fundamental problem and try to develop alternative approaches. there are many approaches to reduce the non-uniqueness. using primary Information is the one of the most promising approach. It is well known that the result of a non-linear least squares Inversion depends highly on the Initial model. for instance, in the one-dimensional case, if we have a priori knowledge (perhaps from a downhole survey) that there are two layers, we can easily estimate a 1D true velocity function from a traveltime curve that indicates two layers. On the contrary, it is very difficult to estimate a true velocity function from the same traveltime curve if we know from other sources that they are actually three layers. Constraints during an Inversion are also very important. Most of the geophysical Inversions need spatial regularization in order to obtain stable results. During the Inversion of the surface seismic refraction method, a constraint that velocity is increasing with depth is generally very effective. in this presentation, we demonstrate the non-uniqueness of the seismic refraction method and the effectiveness of using appropriate primary Information and constraints.