Next: Conclusions
Up: Kerr and Rickett: Long
Previous: Processing flow
Figure 3 shows the impulses reponses derived by
autocorrelation and spectral factorization. The ground-roll cone is
clearly visible with a velocity of about 150 m/s.
Events that may possibly be reflections are visible,
however results are definitely not conclusive.
longimp
Figure 3 Impulse response calculated by (a)
3-D autocorrelation, and (b) spectral factorization. Compare with
Rickett and Claerbout (2000), Figure 3.
Remaining processing problems include how to adaptively remove
monochromatic frequency spikes from the raw data, as noise sources
drift in and out over time. Ground-roll removal is also a problem
since with the 5 m geophone spacing, the 150 m/s velocities are
spatially aliased at all frequencies above 15 Hz.
The other option of imaging reflectors outside the ground-roll window
is restricted by the maximum-offsets of the experiment to events less
than 200 ms two-way traveltime.
Next: Conclusions
Up: Kerr and Rickett: Long
Previous: Processing flow
Stanford Exploration Project
4/29/2001