A particle/collision model of seismic data

Kristopher A. Innanen

It can occasionally be a valuable exercise to take a familiar phenomenon and observe it from a different point of view. It turns out, for instance, that in the right frame of reference, waves in a VSP experiment seem to act like particles, drifting freely or accelerating in a potential, and thereafter colliding, sticking together, and spontaneously disintegrating. In fact, all of the important properties of simple wave phenomena (phase velocities, propagation directions, amplitudes, reflection and transmission coefficients) are correctly captured by speaking of the experiment entirely in terms of a system of colliding particles (with well-defined, though notional, masses, velocities, and momenta). In the same framework, a seismic event, meaning a coherent arrival of wave energy to which we assign a well-defined history of propagation, reflection, and transmission, can be represented in one of two ways. Either a single particle, whose "world-line" is free to move both forwards and backwards in time, or several particles interacting through a specified set of the aforementioned productions and annihilations.