Skip to content

Do you know your seismic velocity?

In seismic data analysis a lot of work goes towards getting a good velocity model of the subsurface.

We talk about the seismic velocity. However, this may lead to some serious misconceptions. One step in seismic processing is velocity analysis. We pick the velocity that adjusts our data for the best value. Usually this is a quite colorful process, as you can see on the image:

Semblance analysis
Seismic Velocity Analysis - GPL Seismic Unix

This image shows which velocity will maximize the local energy over the total energy. This velocity is our so-called stacking velocity or also called normal moveout velocity. Initially, these may sound a little weird, but it’s not that hard actually. The normal moveout or NMO is the delay of arriving reflected energy when we increase the source-receiver distance (called offset). As long as the location of the source and the receiver are coincident, the path of the reflected energy is straight to the reflector and back up to the receiver.

What is stacking?
Stacking is the process of building a combined image of redundant data. In the most basic case this is a simple sum of the energy of our seismic traces.
However, due to the set-up of our seismic experiment, the illumination of a certain reflector will be done with varying source and receiver locations and hence varying offset. Unfortunately, this normal moveout is a non-linear stretch of the traces. This stretch can be corrected for by picking the right apparent velocity of the subsurface. Nevertheless, the velocity we picked is not even close to the real velocity as a rock property.

Stacking velocity and Inversion

The stacking velocity is a tool. It’s simply used to minimize the error, when correcting our seismic traces. However, we can utilize a process called inversion to find the velocity intervals in our seismic data. The easiest process to do this is the Dix inversion. It can get as complicated as a Full Waveform inversion that will take up significant computing power.

Of course the use of a Dix inversion is very limited. It assumes homogenity of the subsurface and availability of RMS velocities, but it’s an inversion that can actually be done with pen and paper.

What velocities are there actally?

We have a wide variety of velocities.

Velocity Usage
Vp vs. Vs Velocities of longitudinally or transversally polarized waves. They’re the two types of body waves we usually look at.
Vnmo The NMO velocity that corrects ideally for the NMO
Vrms The Root Mean Square velocity. An average velocity of the different rocks.
Vstack The stacking velocity minimizes the error of the stack
Vint The interval velocity gives the velocity between two reflectors. Needed for migration; closest to an actual velocity.
Vmean The mean velocity also differs from the other velocities.

All of these velocities have a specific purpose in the seismic world, however these velocities have some very distinct limits as well. These limits often include a phenomenon called anisotropy.

Continue reading "Do you know your seismic velocity?"