Depth conversion of velocity data using velocity analyses

Abstract

Routinely produced velocity analyses are used for a number of purposes in processing and interpretation of seismic data. These analyses have errors due to limitations in techniques and due to a number of assumptions made in their production. The random errors are much greater than the errors in reflection times. These random variations can be separated from the real velocity variations by averaging or smoothing a number of velocity analyses. However, if this is done by simply smoothing average velocity, there is a risk of introducing error by smoothing the variation in average velocity with depth. This can be avoided by smoothing instead the velocity Anomaly. This is the difference at a given time between the measured velocity and the standard velocity taken from a time-velocity curve adopted for the area. An example of the method in the Gippsland Basin shows a marked improvement in accuracy of ties to wells.