This approach suffers from some possible pitfalls.Obtain zero offset section, , by stacking input data. Use to compute smooth reflector dip, , using technique of Fomel (2002). Setk=1,y=y. do while Set Set Set end do_{0}

- In the `` to zone'', consists of
primary events only, and to a (normally) lesser extent, other coherent noise
modes (e.g. locally-converted shear waves, internal multiples). For
, contains both peglegs and primaries, as
well as other possibly strong unmodeled multiple reflection modes, and the
dip of these events is unlikely to be equal.
Previous authors Brown (2002b); Fomel (2001) have developed methods to simultaneously estimate two crossing dips. The problem is inherently nonlinear, and thus highly dependent on initial guesses for the two dips. Assuming weak non-primary/pegleg noise, a possible strategy might be to seed the pegleg dip with primary dip from above.

In current implementation, however, I ignore this problem by setting the dip to zero in the zone. I justify this assumption by noting that as increases, the distance from

*y*to shrinks asmyptotically to zero, so unless the reflector dip is severe (inconsistent with the derivation of equation(3)), this omission will not present a large problem._{0} - Faults and other event discontinuities in are inconsistent with the ``smooth'' mentioned above. The current implementation ignores any possible faulting, although the dip estimation algorithm used can handle steeply folding reflectors.

7/8/2003