A brief synopsis of things you will see later if you keep your mouse out of the way when scrolling!

What you see on a section is almost always a tuned combination of several primary events. Though it may come as a surprise to some, the earth is elastic. When it is hit by a force it bends, then rebounds. Downward traveling seismic waves thus get leggy as they travel. We can argue about the exact shape, but it is a flat fact that several nodes will exist, and that the wavelet will "lap over" multiple reflecting interfaces, as shown below. Now add these primary reflections algebraically in your mind, and notice that  the resulting events do not line up with any bed!  Then - if one interface drops out, the whole damn thing changes.  All who do interpretation should accept this simple seismic reality. The fact is that most do not. This is your intro to tuning.

And it is why inversion is so important to interpretation.  The fact that much AVO work does not take tuning into account is why I don't buy the thesis. The key is to always "think tuning". Numerous examples of increased resolution via inversion are shown on the Power Points you will find inside. How tuning affects AVO is also discussed.

A major seismic purpose is to detail stratigraphy. This example shows what I was able to do in a noise free environment (Qatar gas field). The display  consists of "simulated sonic traces" constructed from my inversion of stacked data.  This is a new technique you probably have not seen before, so pay attention. It is described inside. The well information shows a remarkable match. Take time to study it. Then visualize extending reservoirs using this advanced technique. Jumping too soon to automated interpretation logic misses this all important detail.

But - the input had to be good to get such results, and this is not often the case on land and shallow water data. Again, over-automation hides fatal problems under the rug. Searching for this level of perfection is what drove me to the noise analysis I discuss at length. One's eye has to be trained to see complex patterns.  Once I started seeing noise, it jumped out continually.  So I ask you to look at some noise being removed by non-linear means.

Coherent noise comes in several different forms. Often, as is the case here, there is a central cone that probably contains shear waves, with associate refractions pealing off from this strong energy. Obviously the subject is too complex for this introduction. However, the key in my non-linear work is pattern recognition, using all dimensions available. Once the noise is predicted, it is lifted off, hopefully with no disturbance to the underlying real events. In these two "before and after" pairs, the coherent noise on the field records can be more than 10 times as strong as the reflections. removing it (as shown to the right of each pair) completely changes the reflection clarity.   New work is greatly improving the central cone picture from what is shown here. The sinusoidal problem - If you look closely, you will see places where the main reflection event shows up in the "original" (where it is in phase with the noise). This merging of energy makes resolution very difficult.   When the noise is detected (and gently removed) we see things for the first time. So does the stack, of course.

  Now, if you don't see a left panel, click here to go to the right place