Interpretation Problems in Areas with Complex Structure: Examples Based upon Seismic Forward Modelling and Processing

D. Waltham, K. R. McClay, X. T. Wang

Many standard seismic processing techniques, such as stacking and time migration, perform badly in areas of complex structure. Processing artifacts appear on the final sections which can mislead the unwary interpreter. This poster investigates these problems by looking at synthetic data generated over known structures. The geometric models for the seismic reflectors were based upon laboratory models constructed from sand. These models simulate realistic geometries from many different settings. In this study we have concentrated upon extensional examples which simulate listric and planar fault tectonics. The analogue model geometries were scaled from their laboratory dimensions of, about, 10 cm deep to a scale of, typically, 5 km deep. Seismic P-wave and S-wave velocities ere then assigned to each of the model layers together with densities. Shot records were simulated and the results processed to give final, time migrated, sections. The results, reproduced using many 2-D and 3-D structures, demonstrate that consistent artifacts are produced in nearly all cases. In particular, problems tend to occur beneath major faults leading to a "dead zone" in which virtually no imaging occurs. Under some circumstances this effect can lead to totally incorrect identification of the fault location and geometry. Other important problems result from the velocity distortions of the deeper reflector geometries. These effects, once recognised, can be seen on real seismic sections although they are usually less severe than on our synthetic examples. This difference is a cons quence of the smaller velocity contrasts resulting from burial effects in real cases.

AAPG Search and Discovery Article #91020©1995 AAPG Annual Convention, Houston, Texas, May 5-8, 1995