Seismic traveltime tomography is a nonlinear inverse problem wherein an unknown slowness model is inferred from the observed arrival times of seismic waves. Nonlinearity arises because the raypath connecting a given source and receiver depends on the slowness. Specifically, if L(s) designates a raypath through the slowness model s between two fixed endpoints, then the path integral for traveltime
t(s)=L(s)s dl

is a nonlinear functional of s because it does not, in general, satisfy the superposition condition (i.e., t(s1, + s2) ≠ t(s1), + t(s2) where s1 and s2 are two different slowness models). The tomographic inverse problem can be solved after linearizing the traveltime expression about a known slowness model s0. This linearized expression is usually obtained by appealing to Fermat's principle (e.g., Nolet, 1987). Alternately, the required relation can be rigorously derived via ray-perturbation theory (Snieder and Sambridge, 1992). The purpose of this note is to present a straightforward derivation of the same result by linearizing the eikonal equation for traveltimes. Wenzel (1988) adopts this approach, but his method of proof cannot be generalized to heterogeneous 3-D media. A full 3-D treatment is given here. The proof is remarkably simple, and thus it is quite possible that others have discovered it previously.

This content is PDF only. Please click on the PDF icon to access.

First Page Preview

First page PDF preview