Abstract

In this paper, we are interested in determining the seismic velocity of a shallow under-ground layer from refraction traveltimes measures. We present a study case taken from an experimental seismic survey. The study case is a wide-angle seismic inversion using experimental traveltimes measures and based on ray tracing technique and genetic algorithms. The hypothesis on the velocity distribution, coming from the seismic experiment, makes the computation of some seismic rays expensive in time. We propose to reduce the computations time by introducing a formulation of the inverse problem that avoids such costly rays, hence the inversion becomes feasible. Also we present a sensitivity analysis based on a singular value decomposition of the jacobian of the traveltimes with respect to velocity. We give the relationship between the traveltimes measure errors and the velocity estimation error. We discuss the advantages of this method over the classical one based on the resolution matrix.

Highlights

  • Geological investigations are usually based on seismic surveys that allow obtaining images of under-surface sublayers and interfaces between these layers

  • In this paper, we presented a traveltime inversion seeking to determine the seismic velocity of shallow layer

  • In this special case of shallow layer, the computation of some seismic rays may become very expensive in time

Read more

Summary

INTRODUCTION

Geological investigations are usually based on seismic surveys that allow obtaining images of under-surface sublayers and interfaces between these layers. We consider a class of sensitivity analysis methods called optimal vectors method It is based on computing the singular value decomposition (SVD) of the jacobian of the output data with respect to the input parameters (see [7] and [8]). This method gives a relationship between the level of noise in the traveltime measures and the estimation error of the velocity distribution. We are considering a simple and realistic case of seismic inversion where it is easy to predict intuitively the sensitivity analysis results and to verify the validity of the method

THE FORWARD PROBLEM
THE EIKONAL EQUATION Eikonal equation is derived from the waves equation:
THE SHOOTING METHOD
MODEL SPACE AND DATA SPACE
FRÉCHET DERIVATIVE OF THE TRAVELTIME WITH RESPECT TO THE VELOCITY
DATA ACQUISITION
NUMERICAL RESULTS
CONCLUSION
Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call