next up previous print clean
Next: Does it work? Up: INTERPOLATION WITH P.D.E. DIP Previous: Spatial interpolation with

PDE to reject two dips

You can reject two dips with the operator  
 \begin{displaymath}
(\partial_x-p_1\partial_t)(\partial_x-p_2\partial_t)\end{displaymath} (1)
Finding the two values p1 and p2 is a nonlinear problem that is easy. Let u be the input signal and v be the output signal. Consider  
 \begin{displaymath}
\left(
 {\partial^2 \over \partial x^2}
 + a
 {\partial^2 \o...
 ...\partial t^2}
 \right) \, u(t,x)
 {=}v(t,x) \quad\approx\quad 0\end{displaymath} (2)
Now recognize that (2) (which is a separate equation at each point in the (t,x) space of v(t,x)) is an overdetermined set of linear equations for the two unknowns a and b. It is easy to find a and b which by comparison with equation (1) gives p1 and p2 by the nonlinear but easy equations a=p1+p2 and b=p1p2. The minimum signal required is three seismograms (on which to express $\partial_{xx}$).

To recapitulate, first a simple procedure gives us the required coefficients for a filter that fits two waves to the dataset u(t,x). Second, the same dip filter coefficients can be applied on a mesh in which t and x are interleaved, thus introducing new data locations on which we need data. Third, we interpolate t by any method. Fourth, we find missing traces by minimizing the power in v(t,x).


next up previous print clean
Next: Does it work? Up: INTERPOLATION WITH P.D.E. DIP Previous: Spatial interpolation with
Stanford Exploration Project
1/13/1998