Geophysical data integration and its application to seismic tomography

Conclusions

As mentioned above, geophysical data integration can greatly affect the quality of estimation of attributes by reducing the uncertainty in the model. In this paper, I reviewed the structural similarity and some measurement techniques. The efficiency of these techniques was compared with a simple example, where the physics were not taken into account to emphasize the effect of data-integration techniques. The results suggest that the dip-residual method provides better continuity in estimated structures than does the cross-gradient function. Model estimation results using cross-gradient function shows higher sensitivity, which can be explained by its differentiating nature.

I also compared these methods on a velocity estimation problem, where the physics are also included in the problem statement. I used two different types of auxiliary data with different frequency content to compare the efficiency of the techniques reviewed above. In this case, the cross-gradient function shows a stronger structure-imposing effect. However, the cross-gradient functions are not guaranteed to produce improved results when the frequency contents of the main data field and the auxiliary data are very different. The dip- residual method integrates less structural information by missing some of the anomalies. More examples are suggested to validate the comparison results.

Geophysical data integration and its application to seismic tomography