diffraction-in seismic

[Tingpingbing]

For reflections, we have this curve in the CMP gather.
What will happen for diffraction? How do we interpret these? Why does it have a hyperbolic shape?
What does the diffraction tell us about the structure?
Why does the diffraction apex have low velocity and increase with the flanks?

[GuyM]

What will happen for diffraction?
You get a curve shape on an CDP gather; this can be "flattened" using NMO, but the velocity needed to flatten the diffraction will be higher than the true stacking velocity that flattens the reflections. Not that diffractions and reflections can appear at the same zero-offset two-way-time.

How do we interpret these?
On a stacked, unmigrated (normal ray) section diffractions show you where there is a rapid lateral change in rock properties. This might be a small anomaly (ef volcanic intrusion, gas pocket, carbonate reef, steep-sided feature) or a geological fault.

Why does it have a hyperbolic shape?
Um, that's how waves behave? You can model these with Huygen's Principle, which is a mathematical model of wave behavior we can use to predict wave effects. The energy is being scattered. If you have a coherent event (ie no discontinuity) the diffractions sum to give a reflection.

What does the diffraction tell us about the structure?
Roughly its lateral extent and two way time, and that is is too small to resolve on the seismic data.

Why does the diffraction apex have low velocity and increase with the flanks?
At the apex it will have the same velocity as a reflection, because that's kind of what it is. (A reflection is a series of summed diffraction); the velocity isn't "increasing with the flanks" art all; this is an apparent velocity. The diffraction does not obey the normal mover out equation (which applies to reflections) and so to flatten it requires an NMO correction corresponding to an anonymously high (stacking) velocity, but that is *not* a value that has any geophysical or geological meaning. It's an artefact of the mathematical model being used inappropriately(the NMO equation), not reality.

[Tingpingbing]

Hello again,
Thank you for the explanation.

Do you also know what a ladder effect is for diffraction?

[GuyM]

Not come across that term but it seems to be used when there's complex back-scatter patterns. Back-scatter is another way of saying diffractions.

Where you have a rough surface - for example a seafloor that has a small amount of soft sediment over a steeply dippling, uplifted, harder and layered rock formation you will get a very rough (rugose) surface, that may well scatter a lot more than it reflects.

The same happens on land in the near surface with a large number of small cracks and faults caused by weathering and so on.

I think there's a mathematical reason underpinning that description, but that's all I have come across.

[Tingpingbing]

Thank you