The authors demonstrate how the concept of Point Spread Functions can be successfully applied to time-lapse seismic when the respective acquisition geometries are very different and cannot be closely repeated.
