Conclusions

While f-x prediction generally produces results similar to those of t-x prediction, f-x prediction may create spurious events when parallel events occur in the presence of noise. These spurious events are caused by the long filter length in time of the effective time-domain filter of the f-x filter. The t-x prediction avoids the generation of false events in the presence of strong parallel events by limiting the length of the filter in time. With both f-x and t-x predictions, wavelets may be distorted, since events with wavelets may be considered as closely spaced sets of parallel events. In typical seismic data, this problem is unlikely to be a serious problem, but if a few isolated reflections appear in a field of noise, weak events may be generated that will confuse the interpretation. In the next chapter, a solution to this shortcoming of f-x and t-x prediction will be presented.