previous up next print clean
Next: Velocity analysis at fixed Up: Cole: A drill-bit source Previous: LOCATING SOURCES IN 3-D

VELOCITY ANALYSIS

Rather than selecting a good stacking velocity by trial and error, it would be better to perform a velocity analysis. Velocity analysis proceeds much like the search for sources in 3-D, except now we scan over velocity in addition to scanning over the three spatial dimensions and over time.

The result of velocity analysis is a five-dimensional data volume. This means that the technique is expensive. As an example, using 20 points in x and y, a single depth, 30 velocities, and records containing 4000 time samples and 240 input channels requires roughly 80 minutes of cpu on a Convex C-1. More importantly, the five-dimensional volume means that the result will be difficult to display and interpret. We can easily remove one dimension from the result by performing semblance computations over a single window whose length is that of the entire trace. For sources that should be operating continuously over time, this is a sensible thing to do.

To further simplify interpretation of the results, we can hold one or more parameters fixed while the others are varied. This approach is taken in the following section. I hold x and y fixed to the known location of the drill bit (and other sources) and compute semblance as a function of stacking velocity and source depth.



 
previous up next print clean
Next: Velocity analysis at fixed Up: Cole: A drill-bit source Previous: LOCATING SOURCES IN 3-D
Stanford Exploration Project
12/18/1997